altair.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# TL;DR\n",
    "* Altair introduces sync committees for light client functionality, and reforms to validator rewards/penalties which will have an impact on validator profitability\n",
    "* The variability of total rewards will increase; this particularly impacts solo stakers\n",
    "* Although issuance-neutral in the ideal case, Altair introduces harsher penalties for delayed attestations, and will result in a slight reduction in rewards and therefore net issuance\n",
    "* Incidents on mainnet which result in delays to attestation inclusion will have a much greater impact on rewards than at present\n",
    "* Validators are advised to keep an eye on attestation delay performance; those whose attestations are frequently delayed are likely to be much less profitable after Altair\n",
    "\n",
    "\n",
    "# Introduction\n",
    "In the [last article](https://pintail.xyz/posts/validator-rewards-in-practice/) we looked at the real-world performance of validators on Ethereum's beacon chain — now generally known as the \"consensus layer\". With the beacon chain running smoothly since launch in December 2020 (with only a couple of [minor](https://medium.com/prysmatic-labs/eth2-mainnet-incident-retrospective-f0338814340c) [hiccups](https://blog.lido.fi/lido-20-08-2021-orphaned-blocks-in-ethereum-incident-postmortem/)), many people will have had their attention on the deployment of the [London hard fork](https://blog.ethereum.org/2021/07/15/london-mainnet-announcement/) (including the widely discussed EIP-1559 fee market change) and the coming eth1+eth2 merge, when Ethereum's execution layer will switch to using the beacon chain for consensus, meaning an end to Proof of Work mining.\n",
    "\n",
    "Meanwhile, with less fanfare, consensus client developers have been focused on the beacon chain's first upgrade, known as Altair. This fork will introduce light client functionality and will serve as the first run-through of the process for coordinating a fork on Ethereum's Proof of Stake consensus mechanism. The Altair specification applies some lessons learned since the beacon chain's launch to improve its incentivisation structure and performance, in part by making some alterations to the way rewards and penalties are allocated, and therefore will impact to some degree on validator rewards.\n",
    "\n",
    "In this article, we're going to focus on the economic changes coming Altair. We'll try and understand the likely impact on the validators, using data from mainnet (and some assumptions) to see how validator rewards would have been different, had Altair been active since beacon chain genesis. This will help validators to know what to expect when the fork hits following [upgrades on the Pyrmont and Prater testnets](https://blog.ethereum.org/2021/08/25/finalized-no-28/). Since the identification of [an issue on the Prater testnet](https://hackmd.io/@benjaminion/wnie2_210910#Altair), the mainnet upgrade is expected to occur around mid-October.\n",
    "\n",
    "\n",
    "# Reward Scheme Changes\n",
    "\n",
    "The first change to understand is that in the concept of *base reward* means something slightly different under Altair. The base reward is the basic unit of rewards allocated per epoch. Previously, up to one whole base reward available for performing *each* of the four validator duties (source vote, target vote, head vote, prompt inclusion). However, under Altair we redefine base reward, such that it is the long run average per-epoch reward that would be paid to a perfect validator for fulfilling *all* of its duties. We keep the maximum issuance the same as before, but rather than receiving multiples of the base reward, validators are rewarded with fractions of the base reward for each of their duties.\n",
    "\n",
    "## Reward Weights\n",
    "\n",
    "In addition to redefining the meaning of *base reward*, the weightings allocated to the various duties, and indeed the duties themselves have been altered. The charts below illustrate the \"before\" and \"after\" allocations, assuming perfect validator performance."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x720 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot reward allocation\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "head = 1\n",
    "source = 1\n",
    "target = 1\n",
    "delay = 7/8\n",
    "proposer = 1/8\n",
    "\n",
    "fig, (ax1,ax2) = plt.subplots(1,2,figsize=(16,10))\n",
    "\n",
    "ax1.pie(\n",
    "    [head, source, target, delay, proposer],\n",
    "    labels=['head', 'source', 'target', 'delay', 'proposer'],\n",
    "    autopct='%1.1f%%'\n",
    ")\n",
    "ax1.set_title(\n",
    "    \"Pre-Altair Reward Weights for Validator Duties\"\n",
    ")\n",
    "\n",
    "HEAD_WEIGHT = 14\n",
    "SOURCE_WEIGHT = 14\n",
    "TARGET_WEIGHT = 26\n",
    "SYNC_WEIGHT = 2\n",
    "PROPOSER_WEIGHT = 8\n",
    "WEIGHT_DENOMINATOR = 64\n",
    "\n",
    "ax2.pie(\n",
    "    [HEAD_WEIGHT, SOURCE_WEIGHT, TARGET_WEIGHT, SYNC_WEIGHT, PROPOSER_WEIGHT],\n",
    "    labels=['head', 'source', 'target', 'sync', 'proposer'],\n",
    "    autopct='%1.1f%%'\n",
    ")\n",
    "ax2.set_title(\n",
    "    \"Altair Reward Weights for Validator Duties\"\n",
    ")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Proposer and Delay Rewards\n",
    "\n",
    "From the charts above the first change to notice is that the proposer reward has increased by a factor of four. You may recall that in the pre-Altair spec, we had four equal attestation rewards, but that the fourth reward was split between the attesting validator, who received up to ⅞ of the reward, inversely scaled to the inclusion delay, and the block proposer who would receive ⅛ of the the reward. One consequence of this was that only 3% of validator rewards were allocated to block proposal. As [pointed out](https://github.com/ethereum/eth2.0-specs/issues/2152#issuecomment-747465241) by Danny Ryan shortly after the beacon chain launch, having such a low proportion of validator rewards allocated to block proposal was never the intention of the researchers and is effectively a bug in the spec. This error is corrected in the Altair spec, with block proposers being allocated ⅛ of *total* rewards as originally intended, rather than ⅛ of ¼ of rewards, as was the case in the pre-Altair spec.\n",
    "\n",
    "Meanwhile, the \"delay\" reward has been removed entirely. Instead, the other attestation rewards (head, source and target) are given different inclusion deadlines:\n",
    "* correct head votes are only rewarded if included in the following slot\n",
    "* correct source votes are only rewarded if included within 5 slots (i.e. `integer_squareroot(EPOCH_LENGTH)`)\n",
    "* correct target votes are only rewarded if included within 32 slots (i.e. `EPOCH_LENGTH`)\n",
    "\n",
    "This neatly rewards prompt attestation in a logical way. In particular, the head vote can only help the network to reach consensus on the head of the chain if it is received quickly. The target vote is useful to the network as long as it is included within one epoch, and so validators are rewarded for a correct target vote as long as it is included within 32 slots. The source vote on its own doesn't actually help the chain to reach consensus (but only attestations with a correct source vote can be included at all). Therefore the reward for the source vote is paid if the attestation is included within 5 slots. This deadline — `integer_squareroot(EPOCH_LENGTH)` — was chosen as geometrically half way between the other two rewards. Thus the previous graded rewards scheme for correct head, target and source votes somewhat mimics the \"delay reward\" from the previous version of the spec, in paying more to validators whose attestations are included quickly.\n",
    "\n",
    "Finally, the weightings on the attestation rewards have been changed, with source and head rewards being reduced from $\\frac{16}{64}$ to $\\frac{14}{64}$, and the the target reward being increased from $\\frac{16}{64}$ to $\\frac{26}{64}$. This rebalance reflects the reality that a correct target vote is the most important part of an attestation. As long as the network can come to consensus on the target each epoch, the chain can still finalise. \n",
    "\n",
    "\n",
    "## Sync committees\n",
    "\n",
    "The final difference in the reward scheme is the addition of a new reward for participation in a *sync committee*. This implements the key new feature introduced in Altair which is a mechanism by which light clients can sync with the network. The sync committee is a set of 512 validators which signs every beacon chain header. To ensure that light clients can know who the participants of the sync committee are without keeping the whole beacon chain state themselves, the sync committee rotates relatively infrequently — every 256 epochs or about 1 day.\n",
    "\n",
    "As with block proposal, the membership of the sync committee is a random selection of validators which is made for each sync committee period of 256 epochs. For the duration of the sync committee, those validators can earn the sync committee reward for each slot in which they participate in the sync committee."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "321.0 sync committees are selected each year\n",
      "with a validator set of 200000 validators, on average each validator will be assigned to a sync committee 0.82 times per year\n"
     ]
    }
   ],
   "source": [
    "# calculate number of sync committees and expected number of committee selections per validator\n",
    "\n",
    "SECONDS_PER_SLOT = 12\n",
    "SLOTS_PER_EPOCH = 32\n",
    "EPOCHS_PER_COMMITTEE = 256\n",
    "COMMITTEE_SIZE = 512\n",
    "SECONDS_PER_YEAR = 31556952\n",
    "\n",
    "seconds_per_committee = SECONDS_PER_SLOT * SLOTS_PER_EPOCH * EPOCHS_PER_COMMITTEE\n",
    "committees_per_year = SECONDS_PER_YEAR / seconds_per_committee\n",
    "\n",
    "print(f\"{committees_per_year:.1f} sync committees are selected each year\")\n",
    "\n",
    "NUM_VALIDATORS = 200000\n",
    "\n",
    "expected_committee_selections_per_year = committees_per_year * COMMITTEE_SIZE / NUM_VALIDATORS\n",
    "\n",
    "print(f\"with a validator set of {NUM_VALIDATORS} validators, on average each validator will be assigned \"\n",
    "      f\"to a sync committee {expected_committee_selections_per_year:.2f} times per year\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In a similar manner to the way we looked at the variation in proposal duties in a [previous article on beacon chain rewards](https://pintail.xyz/posts/beacon-chain-validator-rewards/#block-rewards), we can use the binomial distribution to look at the variation in how many times a validator can expect to be selected to participate in a sync committee. For the below calculation, we'll assume the total number of validators is 200,000."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.4391785090173583\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot pdf\n",
    "\n",
    "from scipy.stats import binom\n",
    "\n",
    "SECONDS_PER_YEAR = 31556952\n",
    "SECONDS_PER_SLOT = 12\n",
    "SLOTS_PER_EPOCH = 32\n",
    "COMMITTEE_EPOCHS = 256\n",
    "NUM_VALIDATORS = 200000\n",
    "COMMITTEE_VALIDATORS = 512\n",
    "epochs_per_year = SECONDS_PER_YEAR / (SECONDS_PER_SLOT * SLOTS_PER_EPOCH)\n",
    "\n",
    "x = [el for el in range(7)]\n",
    "p_selection = COMMITTEE_VALIDATORS / NUM_VALIDATORS\n",
    "committees_per_year = epochs_per_year / COMMITTEE_EPOCHS\n",
    "y = binom.pmf(x, committees_per_year, p_selection)\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(12, 8))\n",
    "ax.bar(x, y)\n",
    "ax.set_xlim(xmin=-0.5)\n",
    "ax.set_ylim(ymin=0)\n",
    "ax.set_title('Probability mass function (200,000 validators) — number of sync committees per year')\n",
    "ax.set_xlabel('Number of sync committee selections in a year')\n",
    "ax.set_ylabel('Proportion of validators')\n",
    "\n",
    "print(y[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So with 200,000 active validators, almost half of the validators will not be selected for a single beacon committee over the course of a year. If the size of the validator set continues to increase, the probability of being selected for a validator committee will drop even lower.\n",
    "\n",
    "# Modelling Perfect Participation\n",
    "\n",
    "With this in mind, let's start by modelling the distribution of possible annual rewards available, assuming all validators execute their duties perfectly. Recall that in a pre-Altair world, even under perfect participation, there was variability in the validator rewards due to the random allocation of proposer duties. After Altair, the value of block proposals increases by a factor of 4 (from 3.1% of total issuance up to 12.5%), so the variability in validator rewards will accordingly increase. The introduction of sync committees is an additional source of variation in validator rewards.\n",
    "\n",
    "Since sync committee rewards occur randomly and independently of proposer rewards, we can calculate the distribution of total annual rewards by calculating every possible combination of `num_proposer_duties` and `num_sync_committees` in a year, multiplying together the probabilities from each distribution and summing the reward amounts. We can then compare this distribution to the simple binomial distribution which describes the variation in validator rewards pre-Altair."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "code_folding": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Pre-altair annual reward statistics (ETH)\n",
      "-----------------------------------------\n",
      "             median: 2.1031\n",
      "               mean: 2.1038\n",
      " standard deviation: 0.0181\n",
      "interquartile range: 0.0200\n",
      "\n",
      "Altair annual reward statistics (ETH)\n",
      "-------------------------------------\n",
      "             median: 2.0951\n",
      "               mean: 2.1038\n",
      " standard deviation: 0.1025\n",
      "interquartile range: 0.1400\n",
      "\n",
      "relative spread: 5.7 (standard deviation) / 7.0 (interquartile range)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtcAAAJcCAYAAAAy+YhGAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAABeL0lEQVR4nO3deZgsZXn38e9PVpVVwIX1oKJ4XFA5QFRU3EFFNKKguOtLTMQlRiMag4iJwagkGnEhLiCKaMCFCIorooLCAVk8IOaIIIsosoOCHLjfP6oG+rSz9MxUz0xzvp/rmmu6tqfufrqq+u6nnqpKVSFJkiRp9u423wFIkiRJdxUm15IkSVJHTK4lSZKkjphcS5IkSR0xuZYkSZI6YnItSZIkdcTkWiMpyYFJPjeL5Zcl2aW7iOZPkkrywPmOY6aS7JLk0iGVvU+Sb3VY3h3bzWy3wXHKfkeST3ZVXl/ZmyT5RZK7D6P8UZbkFUl+1DN8Y5L7DzLvqiKNzyS5Jslp8x3PbA16/J9sW+gghm8keXkH5bw+yfu6iEndMbnWjCW5KMmf2gPQ75IcnmSdjtdxeJIVSe43yTyL2gRz9UHLraqHVtVJM4jnwHZdO0132fmQ5KQkN7ef0R+SfHmyuhwl7bbx5yQ3tH8/T/JvSdYfm6eqPl9VTx+wrH+Zar6ZbjfjrO8vflBU1Xur6jWzLXsC+wOHV9Wf2vV/IMn/tfX2iyQv64vvkUnOSPLH9v8je6YlyfuSXNX+vS9JJlpxkhcnuTjJTUm+muRePdPuleQr7bSLk7x40GWHparWqaoLZ1tO1z++5tnOwNOAzatqx5kWMh8/Tsbbtwfdj4e5LVTVblV1xGzLBv4b2CfJvTsoSx0xudZs7V5V6wCPBpYA7+yfYTpJb99y9wSeD1wHvGQ2QU5zvePG2yYQLwOubv+Piv3az+iBwDrAB+YrkJluC5P496paF9gEeCXwV8CP222nM0OIe84kWQt4OdD75X4TsDuwfjvtQ0ke286/JvC1dv4NgSOAr7XjAfYFngtsBzyiLedvJlj3Q4FPAC8F7gP8EfhozyyHAn9up+0DfKxdZpBl79IWyjbXxrEVcFFV3TTf8UxHktXmO4Zhq6qbgW8wWt9Jd31V5Z9/M/oDLgKe2jP8fuDr7esCXgf8H/DrdtyzgbOAa4FTgEdMUf7LgEuANwI/75t2IPC59vVv2vXd2P49BngA8D3gKuAPwOeBDcaLvS3rGJpk4nrgNRPE8wTgTzRJwFXAmj3TXgH8iCZxvQb4NbBbz/STgPcAPwZuAL4FbNxO2wW4dKK6BXYETm3r7bfAR/rWXcADJ4j5pN73A/wdsKxneFvg2zQ/GC4AXtiO37pd393a4f8Gft+z3JHAm9rXrwTOb9/XhcDf9My3C3Ap8Dbgina5uwOHt/V0HvDW3vffzntZW94FwFMmeG+HA//SN27dto726/1c2tcB/gP4ffs5nws8jCZZvJUmybsR+N+ez+BtwDnALcDqjL/dfLGN9Uxgu4k+l7F4gXvSbEe3c+c2uyk923Q7/3OAZe3ncBLwkL7t4y1tbNe1Maw9yXa7fIp97TjgH9rXT2/rPz3TfwPs2r4+Bdi3Z9qrgZ9MUO57gaN6hh/Q1vO6bT38GXhQ33Z18FTLjrOejwEf6Bv3NeDN7ev9gV+1n9N5wPP6993xPjdgo7ZurgdOo9mHe+f9EM0x6nrgDODx7fhd21hvbT/fs9vxm7blXQ0sB/5f3zFtpeMQzb6/tB3+HXDIZJ/jOGVNtG1uChwLXElzrHrDJHH8DXAzcFv7Xt7dzjfh8RzYAvhyW/5VNMesh/SVc+0kx6x/a+v7+vZzvFfP9P+hOZZcB5wMPLRvH/sYcALND8jJ9u2x/Xg14B0928cZwBbjbAuHAx+nOV7eAPwA2GoW28JJtMdmmobOdwIX0xyfPgus305b1Mbxcpr98A/AP/XV2T7A9wfZNvybmz9brtWJJFsAzwR+1jP6ucBOwOIkjwI+TXOg3oimReq4tlVtIi8HvgAcDWybZPsJ5ntC+3+Dak7jnUqTSP0bzZfIQ2gO9gdOsq49aL5QNqBJxCeK53+BL7XDu/dN34kmGdwY+HfgU32ny19Mk4jeG1iTJjkaxG3A37flPgZ4Ck2SPC1JNgL+muZLfezMwLeBo9qY9gY+mmRxVf2a5kviUe3iTwBuTPKQdviJNF8u0HwZPBtYr31//5Hk0T2rvi9wL5rWr32Bd9EkSg8AnkFTr2MxPhjYD9ihmhbpZ9B8EQ6kqm5o39Pjx5n89PZ9PIimxfaFwFVVdRjNZ/7v7fbT+7m+CHgWzba1Ypwy96D5sr8XTT1+NckaU8R4E7AbcHm7vnWq6vLeeZI8iGbbfxNNq/wJwP/2tB7Txr8rzQ+hR9AkieN5OM12Oa40/bB3oEnkAR4KnFPtt3brnHb82PSze6ad3TOt30rzVtWvaBPq9m9FVf1ygrImW7bfF4C9xva3JBvSfN5Ht9N/RbNNrA+8G/jcgN2jDqVJCO8HvKr963U68Eju/Pz/J8naVfVNmh8HX2w/3+3a+Y+m+bG5KbAn8N4kT+4pr/849CHgQ1W1Hs3+8iUGN+62meRuNMexs4HNaI4nb0ryjAni+CzwWuDU9r28a7Ljedta/HWaRHFRu46jq+r8vnI2mCT2l9HU9f2AFcCHe6Z9A9iG5ph1Jn95vH4x8K80P+A+y8T79pg30+znz6Q5hr2K5izJePah+YG1Mc0Pi951T3db6PWK9u9JwP1pzjB+pG+enYEH03xeB/Qci6Fp3BivXM0Tk2vN1leTXEvTavsDmoPImH+rqqur6ee5L/CJqvppVd1WTV+zW2hO4/+FJFvSHGiOqqrfAd9lGqe9qmp5VX27qm6pqiuBQ2gSwomcWlVfrarb23j747kH8II2nltpvnj647m4qv67qm6jOZV+P5rT2WM+U1W/bMv/Es2BeJD3ckZV/aSqVlTVRTRfZJO9l34fTnIdTYvHxsDr2/HPpjnV+5m27J/RtGa9oJ3+A+CJSe7bDh/TDm9N8yV0dhvf8VX1q2r8gKZVvje5vR14V/tZ/IkmKfzXdtu4hJW/OG8D1qL5QbZGVV3UJlXTcTnNF1y/W2m+cLelaZU9v6p+O0VZH66qS8bbJlpnVNUx7TZxCLA2E2zT07QXcHy7Dd9Kc0bk7sBj+2K7vKqupkmWHjlBWRvQtLRN5OM0n+WJ7fA6NK2Cva6jqbvxpl8HrDNBv+vJylqH5gfcoOvpn97rhzSte2Pb3Z40+/TlAFX1P21d3V5VX6Q5ozZp3+E2SXw+cEBV3VRVP6fZr+9QVZ+rqqva/eeDNNvugycobwvgccDbqurmqjoL+CQrH0f6j0O3Ag9MsnFV3VhVP5ks5j4TbZs7AJtU1UFV9edq+hT/N82P64ni6DfZ8XxHmh8Pb23r7eaqmm4/6yOr6uftD9F/Bl441sWjqj5dVTdU1S00DSbb9V5nAXytqn7cxn7zAOt6DfDOqrqgPYadXVVXTTDv8VV1crvufwIe036u09oWxrEPzVmJC6vqRuDtwN59XYPeXVV/qqqzafbX3mT6BpofjlogTK41W8+tqg2qaquq+ru+A/ElPa+3Av4hybVjfzStyZumuaPDje3fN9r5Xwqc334BQdNC8OKpWgXHJLlPkqOTXJbkeppTnBtPssglk0wDeB5NC8oJPfHslmSTnnmuGHtRVWMtH+uMN52mZWSgiz+TPCjJ15Nc0b6X9zL5e+n3hqpan6Z1c0Ng83b8VsBOfZ/JPjQtzdAk17vQtPaeTHMa84nt3w+r6vY2vt2S/CTJ1W0Zz+yL78q+L7lNWbm+Lx57UVXLaVprDwR+336Gm07jvULTUnZ1/8iq+h5Na9ChbdmHJVlvirKm2i7umN7Wx1ir5Gxtysr1cnu7rs165hl0e7qG8RNSkryfpmvMC6vuaKm+kebHU6/1uDNB75++HnBjz/K9Jitruuvpn36Hdt1H07RAQtN6eUerYpKXJTmrZzt/GFPvQ5vQdAUad1tty31LkvOTXNeWu/4k5W4KXF3N2ZXe8no/0/7t7dU0LfW/SHJ6kmdPEXOvibbNrWiOu737/TtYuSFgqu1+wuN5+//iGv9Mz7Rjp6mjNYCNk6yW5OAkv2qPhRe182w8wbKD2ILmzMa04mqT4Ktp9/dpbgv9Vtrf29ers/JnMtn+vi5/+UNU88jkWsPU+2V7CU1r5QY9f/eoqi9Uc0eHsdPju7Xzvwy4f5tQXkHT8rIxTeI22XrGvLcd//BqTqm+hKaryCCxjuflNAez37Tx/A/NAf/Fky41mJuAe4wNtC00vUn7x4BfANu07+UdTP5exlVV59L0+T20bWW8BPhB32eyTlX9bbvID2haAndpX/+IpuXtji4hbbeeY2laVu9TzaneE/ri66/b39J8oY3Zsi/Oo6pqZ5ov8AIGvs1UmrvVPJWmJXO8OvhwVW0PLKZJWt46QYwTxd7vjvfRnm7fnKblHJovwHv0zHvfntdTlXs5zfsfKzvtui6bYrnxnMM4XSmSvJume8rTq6q3BXkZ8Ii+luhHcGe3kWWs3Gq2Xc+0fivNm+a2ZmsBv2z/Vk+yzQRlTbbseL4A7JlkK5ouWse2y21F0zK7H7BRu43+nKn3oStpflCPu60meTzwjzRnYjZsy72up9z+z/hy4F5Jen/obMnKn+lKy1TV/1XVi2i6QLwPOCaDX6w70bZ5Cc11ML37/bpV1XtsnWr7nPB43k7bcoILMqcq9y9ip6mjW2nOvL2YpsvKU2mS10Vjb3GSdQzyXh4w3bjaY829gMtnsC30W2l/p3nPK2j62Q/iIazcVUvzzORac+W/gdcm2SmNeyZ5Vt8XDQBJxi5I3JHmVPcjaVqajmL8riFX0nQ96L0f6bo0LV/XJdmMO5OoaWuXfwpNN4qxeLaj+bLr4grtXwJrt/WxBs2FLb190delOX1+Y5Jtgb8dp4xBHUHTGvIcmn6RD0ry0rYv5hpJdhjry1dV/0dz4d1LaJLwsYuqns+d/a3XbGO9EliRZDeavq6T+RLw9iQbJtmcO7upkOTBSZ7cJu03c+eFf5Nq+3puD3yVpqX2M+PMs0O7/a1B84Pm5p6yf8fK28+gtk/y120i8SaaU+Njp+7PojnbslqSXVm5K8/vgI36Tmf3+hLwrCRPaeP9h7bsU2YQ42nABu12DECSt9MkKk8d5xT4STTdc97Q1ut+7fjvtf8/C7w5yWbtWYV/oLnYa6zsi5K8oh38PLB7kse3SeFBwJfb0/o30Vz0dlB7PHgcTeJ05FTLjvcmq+nW9AearhYnVtW17aR70iQ3V7bxvZLmeDKparp3fRk4MMk9kiym5/oAmv1yRVvu6kkOYOWW9t8Bi9rElmq6QJ0C/FuStZM8gqZlesLb9SV5SZJN2pbnsfczdsaot57HM9G2eRpwQ5K3Jbl7u30+LMkOU9VJj8mO56fR/IA+uB2/dvvZjtXJ5ln52oHxvCTJ4jTd8Q4Cjmk/j3Xb93EVzQ/X905Sxpip9u1PAu9Jsk37Xh6R5vqU8Twzyc5t/O+huZD3Eqa5LYzjC8DfJ9m6TdrH+mgP2vr/RJq+6FogTK41J6pqKfD/aE7LX0NzUd0rJpj95TT95s6tqivG/mgu7nl2+u5123bB+FeaW7Bdm+SvaC5aejRN68HxNF+SM/VS4Kyq+lZfPB+maeGb8ot6MlV1Hc0Fip+kacW6ieYU7pi30CRCN9B8qX1xFuv6M009/nObpDydpq/l5TSnHd/Hyon9D2gu+rukZzg0FxKNXUD4Bppk8Jo2zuOmCOPdNKc9f03TP/vInmlrAQfTJElX0LTYvX2Ssv4xyQ00X7afpblK/7E1/i3D1qOpv2va9V9Fc4cbgE/R9PO+NslXp4i/19do+kdfQ7Od/HU1fVyhucvN7jRJ0T40iT8AVfULmi/UC9t1rtSVpKouoPlR8180dbE7zW0v/zyN2MbK+jNN8tt7O8v30rSOLc+dXbLe0TP/c2l+OF5Lc4HXc3vW/QmaPt7n0rQAH9+OG7uN30a0PzCqahnNRWyfp7nwdV1Wvhj372j6kv++rY+/bZcZZNnxHEXTqnlUz/s/D/ggzR13fkdzgeePpyhnzH40Z6yuoKnD3h9tJwLfpPlxfDHNj7XeLgn/0/6/KsmZ7esX0bS2Xg58heZahO9Msv5dgWVJbqTZb/euqj/11/MExt022yR1rKHg19z5g2TgPruTHc/b8nenufXnb2iOZXu1i36P5ozEFUn+MMkqjqSp7yto+oq/oR3/WZq6vozmri+D9EGfat8+hOb49S2aRoxP0WyT4zmK5oLsq4HtuXOfmsm20OvTNO/5ZJrP5GZ6Gh0mk2RtmjO6XdwzWx3J+N3kJEl3FWmuDfgh8Kia+OLMLtazM/C6tiuDhmSqek5yIM0t5Obs+QBdSXISzS0ph/K00plKcjjNLUP/4lkO8ynJ62luHfiP8x2L7rQgblIvSRqeau6Ys+0crOdHNH3zNUTWs8ZU1X/Ndwz6S3YLkSRJkjpitxBJkiSpI7ZcS5IkSR25y/S53njjjWvRokXzHYYkSZLu4s4444w/VNUm4027yyTXixYtYunSpfMdhiRJku7iklw80TS7hUiSJEkdMbmWJEmSOmJyLUmSJHXE5FqSJEnqiMm1JEmS1BGTa0mSJKkjJteSJElSR0yuJUmSpI6YXEuSJEkdMbmWJEmSOnKXefy5JGlwi/Y/fmhlX3Tws4ZWtiQtdLZcS5IkSR0xuZYkSZI6YnItSZIkdcTkWpIkSeqIybUkSZLUEZNrSZIkqSMm15IkSVJHTK4lSZKkjphcS5IkSR0xuZYkSZI6YnItSZIkdcTkWpIkSeqIybUkSZLUEZNrSZIkqSMm15IkSVJH5iW5TrJrkguSLE+y/yTzPT9JJVkyl/FJkiRJMzHnyXWS1YBDgd2AxcCLkiweZ751gTcCP53bCCVJkqSZmY+W6x2B5VV1YVX9GTga2GOc+d4DvA+4eS6DkyRJkmZqPpLrzYBLeoYvbcfdIcmjgS2q6vjJCkqyb5KlSZZeeeWV3UcqSZIkTcOCu6Axyd2AQ4B/mGreqjqsqpZU1ZJNNtlk+MFJkiRJk1h9HtZ5GbBFz/Dm7bgx6wIPA05KAnBf4Lgkz6mqpXMWpSQtMIv2n/RkniRpAZiPluvTgW2SbJ1kTWBv4LixiVV1XVVtXFWLqmoR8BPAxFqSJEkL3pwn11W1AtgPOBE4H/hSVS1LclCS58x1PJIkSVJX5qNbCFV1AnBC37gDJph3l7mISZIkSZqtBXdBoyRJkjSqTK4lSZKkjphcS5IkSR0xuZYkSZI6YnItSZIkdcTkWpIkSeqIybUkSZLUEZNrSZIkqSMm15IkSVJHTK4lSZKkjphcS5IkSR0xuZYkSZI6YnItSZIkdcTkWpIkSeqIybUkSZLUEZNrSZIkqSMm15IkSVJHTK4lSZKkjphcS5IkSR0xuZYkSZI6YnItSZIkdcTkWpIkSeqIybUkSZLUEZNrSZIkqSMm15IkSVJHTK4lSZKkjphcS5IkSR0xuZYkSZI6YnItSZIkdcTkWpIkSepIZ8l1krslWa+r8iRJkqRRM6vkOslRSdZLck/g58B5Sd7aTWiSJEnSaJlty/XiqroeeC7wDWBr4KWzDUqSJEkaRbNNrtdIsgZNcn1cVd0K1KyjkiRJkkbQbJPrjwMXAfcETk6yFXD9bIOSJEmSRtHqM10wyd2A31XVZj3jfgM8qYvAJEmSpFEz45brqrod+Me+cVVVK2YdlSRJkjSCZtst5DtJ3pJkiyT3GvvrJDJJkiRpxMy4W0hrr/b/63rGFXD/WZYrSZIkjZxZJddVtXVXgUiSJEmjblbJdXsbvr8FntCOOgn4RHtLPkmSJGmVMttuIR8D1gA+2g6/tB33mlmWK0mSJI2c2SbXO1TVdj3D30ty9izLlCRJkkbSbO8WcluSB4wNJLk/cNssy5QkSZJG0mxbrt8KfD/JhUCArYBXzToqSZIkaQTNNrn+EbAN8OB2+IJBFkqyK/AhYDXgk1V1cN/019Lc3u824EZg36o6b5axSpIkSUM1224hp1bVLVV1Tvt3C3DqZAskWQ04FNgNWAy8KMnivtmOqqqHV9UjgX8HDpllnJIkSdLQzajlOsl9gc2Auyd5FE2XEID1gHtMsfiOwPKqurAt62hgD+COlumqur5n/nvSPJhGkiRJWtBm2i3kGcArgM1ZuVX5BuAdUyy7GXBJz/ClwE79MyV5HfBmYE3gyeMVlGRfYF+ALbfccrDIJUmSpCGZUXJdVUcARyR5flUd23FMY+s4FDg0yYuBdwIvH2eew4DDAJYsWWLrtiRJkubVbB9/fmySZwEPBdbuGX/QJItdBmzRM7x5O24iR9M8mEaSJEla0GZ1QWOSjwN7Aa+n6Xf9Aprb8U3mdGCbJFsnWRPYGziur9xtegafBfzfbOKUJEmS5sJsb8X32Kp6RJJzqurdST4IfGOyBapqRZL9gBNpbsX36apaluQgYGlVHQfsl+SpwK3ANYzTJUSSJElaaGabXP+p/f/HJJsCVwH3m2qhqjoBOKFv3AE9r984y7gkSZKkOTfb5PrrSTYA3g+cSXPLvE/ONihJkiRpFM32gsb3tC+PTfJ1YO2qum72YUmSJEmjZ6YPkfnrSaZRVV+eeUiSJEnSaJppy/Xu7f97A48FvtcOPwk4BTC5liRJ0ipnpg+ReSVAkm8Bi6vqt+3w/YDDO4tOkiRJGiGzus81sMVYYt36HeBzyCVJkrRKmu3dQr6b5ETgC+3wXsB3ZlmmJEmSNJJme7eQ/dqLGx/fjjqsqr4y+7AkSZKk0TPbluuxO4N4AaMkSZJWeTO9Fd+PqmrnJDfQPDjmjklAVdV6nUQnSZIkjZCZ3i1k5/b/ut2GI0mSJI2umbZc32uy6VV19czCkSRJkkbXTPtcn0HTHSTjTCvg/jOOSJIkSRpRM+0WsnXXgUiSJEmjbtZ3C0myIbANsPbYuKo6ebblSpIkSaNmVsl1ktcAbwQ2B84C/go4FXjyrCOTJEmSRsxsH3/+RmAH4OKqehLwKODa2QYlSZIkjaLZJtc3V9XNAEnWqqpfAA+efViSJEnS6Jltn+tLk2wAfBX4dpJrgItnG5QkSZI0imaVXFfV89qXByb5PrA+8M1ZRyVJkiSNoNle0Phh4OiqOqWqftBRTJIkSdJImm2f6zOAdyb5VZIPJFnSRVCSJEnSKJpVcl1VR1TVM2nuGHIB8L4k/9dJZJIkSdKImW3L9ZgHAtsCWwG/6KhMSZIkaaTMKrlO8u9tS/VBwLnAkqravZPIJEmSpBEz21vx/Qp4TFX9oYtgJEmSpFE221vxfaKrQCRJkqRR11Wfa0mSJGmVN6PkOsnWXQciSZIkjbqZtlwfA5Dkux3GIkmSJI20mfa5vluSdwAPSvLm/olVdcjswpIkSZJGz0xbrvcGbqNJztcd50+SJEla5cyo5bqqxp7GeE5VfaPjmCRJkqSRNNu7hZyS5JAkS9u/DyZZv5PIJEmSpBEz2+T608ANwAvbv+uBz8w2KEmSJGkUzfYJjQ+oquf3DL87yVmzLFOSJEkaSbNtuf5Tkp3HBpI8DvjTLMuUJEmSRtJsW65fC3y2p5/1NcDLZ1mmJEmSNJJmlVxX1dnAdknWa4ev7yQqSZIkaQTNtuUaMKmWJEmSYPZ9riVJkiS1TK4lSZKkjsy6W0iSxwKLesuqqs/OtlxJkiRp1MwquU5yJPAA4CzgtnZ0ASbXkiRJWuXMtuV6CbC4qqqLYCRJkqRRNts+1z8H7jvdhZLsmuSCJMuT7D/O9DcnOS/JOUm+m2SrWcYpSZIkDd1sW643Bs5Lchpwy9jIqnrORAskWQ04FHgacClwepLjquq8ntl+Biypqj8m+Vvg34G9ZhmrJEmSNFSzTa4PnMEyOwLLq+pCgCRHA3sAdyTXVfX9nvl/ArxkFjFKkiRJc2JW3UKq6gfAL4B127/z23GT2Qy4pGf40nbcRF4NfGO8CUn2TbI0ydIrr7xy8MAlSZKkIZhVcp3khcBpwAuAFwI/TbJnF4G15b+E5qLJ9483vaoOq6olVbVkk0026Wq1kiRJ0ozMtlvIPwE7VNXvAZJsAnwHOGaSZS4DtugZ3rwdt5IkT23Lf2JV3dI/XZIkSVpoZnu3kLuNJdatqwYo83RgmyRbJ1kT2Bs4rneGJI8CPgE8p698SZIkacGabcv1N5OcCHyhHd4LOGGyBapqRZL9gBOB1YBPV9WyJAcBS6vqOJpuIOsA/5ME4DeT3YFEkiRJWghmlVxX1VuTPB94XDvqsKr6ygDLnUBfEl5VB/S8fups4pIkSZLmw2xbrqmqY4FjO4hFkiRJGmkzSq6T/Kiqdk5yA9D76PMAVVXrdRKdJEmSNEJmlFxX1c7t/3W7DUeSJEkaXbO9z/WRg4yTJEmSVgWzvRXfQ3sHkqwObD/LMiVJkqSRNKPkOsnb2/7Wj0hyfft3A/A74GudRihJkiSNiBkl11X1b8D6wGerar32b92q2qiq3t5tiJIkSdJomHG3kKq6Hdihw1gkSZKkkTbbPtdnJjHBliRJkpj9Q2R2AvZJcjFwE3fe5/oRs45MkiRJGjGzTa6f0UkUkiRJ0l3ArLqFVNXFwAbA7u3fBu04SZIkaZUz24fIvBH4PHDv9u9zSV7fRWCSJEnSqJltt5BXAztV1U0ASd4HnAr812wDkyRJkkbNbO8WEuC2nuHb2nGSJEnSKme2LdefAX6a5Cs0SfUewKdmHZUkSZI0gmaVXFfVIUlOAnYGCnhlVf2si8AkSZKkUTPbbiFj0vdfkiRJWuXM9m4hBwBHABsCGwOfSfLOLgKTJEmSRs1s+1zvA2xXVTcDJDkYOAv4l1mWK0mSJI2c2XYLuRxYu2d4LeCyWZYpSZIkjaTZtlxfByxL8m2aCxqfBpyW5MMAVfWGWZYvSZIkjYzZJtdfaf/GnDTL8iRJkqSRNdtb8R2RZE3gQe2oC6rq1tmHJUmSJI2eWSXXSXahuVvIRTS34dsiycur6uRZRyZJkiSNmNl2C/kg8PSqugAgyYOALwDbzzYwSZIkadTM9m4ha4wl1gBV9UtgjVmWKUmSJI2k2bZcn5Hkk8Dn2uF9gKWzLFOSJEkaSbNNrl8LvA4Yu+XeD4GPzrJMSZIkaSTNOLlOshpwdlVtCxzSXUiSJEnSaJpxn+uqug24IMmWHcYjSZIkjazZdgvZkOYJjacBN42NrKrnzLJcSZIkaeTMNrn+506ikCRJku4CZpRcJ1mb5mLGBwLnAp+qqhVdBiZJkiSNmpn2uT4CWEKTWO9G8zAZSZIkaZU2024hi6vq4QBJPgWc1l1IkiRJ0miaacv1rWMv7A4iSZIkNWbacr1dkuvb1wHu3g4HqKpar5PoJEmSpBEyo+S6qlbrOhBJkiRp1M34ITKSJEmSVmZyLUmSJHXE5FqSJEnqiMm1JEmS1JHZPv5ckqSVLNr/+M7KuujgZ3VWliTNBVuuJUmSpI7MS3KdZNckFyRZnmT/caY/IcmZSVYk2XM+YpQkSZKma86T6ySrAYcCuwGLgRclWdw322+AVwBHzW10kiRJ0szNR5/rHYHlVXUhQJKjgT2A88ZmqKqL2mm3z0N8kiRJ0ozMR7eQzYBLeoYvbcdNW5J9kyxNsvTKK6/sJDhJkiRppkb6gsaqOqyqllTVkk022WS+w5EkSdIqbj6S68uALXqGN2/HSZIkSSNtPpLr04FtkmydZE1gb+C4eYhDkiRJ6tScJ9dVtQLYDzgROB/4UlUtS3JQkucAJNkhyaXAC4BPJFk213FKkiRJ0zUvT2isqhOAE/rGHdDz+nSa7iKSJEnSyBjpCxolSZKkhcTkWpIkSeqIybUkSZLUEZNrSZIkqSMm15IkSVJHTK4lSZKkjphcS5IkSR0xuZYkSZI6YnItSZIkdcTkWpIkSeqIybUkSZLUEZNrSZIkqSMm15IkSVJHTK4lSZKkjphcS5IkSR0xuZYkSZI6YnItSZIkdcTkWpIkSeqIybUkSZLUEZNrSZIkqSMm15IkSVJHVp/vACTprmzR/sfPdwiSpDlky7UkSZLUEZNrSZIkqSMm15IkSVJHTK4lSZKkjphcS5IkSR0xuZYkSZI6YnItSZIkdcTkWpIkSeqIybUkSZLUEZNrSZIkqSMm15IkSVJHTK4lSZKkjphcS5IkSR0xuZYkSZI6YnItSZIkdcTkWpIkSeqIybUkSZLUEZNrSZIkqSMm15IkSVJHTK4lSZKkjphcS5IkSR0xuZYkSZI6svp8ByBJ0kQW7X98Z2VddPCzOitLkiYyLy3XSXZNckGS5Un2H2f6Wkm+2E7/aZJF8xCmJEmSNC1znlwnWQ04FNgNWAy8KMnivtleDVxTVQ8E/gN439xGKUmSJE3ffLRc7wgsr6oLq+rPwNHAHn3z7AEc0b4+BnhKksxhjJIkSdK0zUef682AS3qGLwV2mmieqlqR5DpgI+APvTMl2RfYtx28MckFQ4l4ahvTF5smZF1Nj/U1OOtqela5+srMz4GucnU1S9bX4Kyr6VlI9bXVRBNG+oLGqjoMOGy+40iytKqWzHcco8C6mh7ra3DW1fRYX4OzrqbH+hqcdTU9o1Jf89Et5DJgi57hzdtx486TZHVgfeCqOYlOkiRJmqH5SK5PB7ZJsnWSNYG9geP65jkOeHn7ek/ge1VVcxijJEmSNG1z3i2k7UO9H3AisBrw6apaluQgYGlVHQd8CjgyyXLgapoEfCGb964pI8S6mh7ra3DW1fRYX4OzrqbH+hqcdTU9I1FfsUFYkiRJ6oaPP5ckSZI6YnItSZIkdcTkekBJPp3k90l+PsH0tyY5q/37eZLbktxrruNcKAaor/WT/G+Ss5MsS/LKuY5xoRigrjZM8pUk5yQ5LcnD5jrGhSLJFkm+n+S8drt54zjzJMmHkyxv6+zR8xHrQjBgfW2b5NQktyR5y3zEuRAMWFf7tNvUuUlOSbLdfMS6EAxYX3u09XVWkqVJdp6PWOfbIHXVM+8OSVYk2XMuY1xIBty2dklyXU/edcB8xDoR+1wPKMkTgBuBz1bVpMlNkt2Bv6+qJ89JcAvQVPWV5B3A+lX1tiSbABcA922f2rlKGaCu3g/cWFXvTrItcGhVPWWu41wIktwPuF9VnZlkXeAM4LlVdV7PPM8EXg88k+YBVR+qqv4HVa0SBqyve9M8DOG5wDVV9YF5CXaeDVhXjwXOr6prkuwGHOi2NWl9rQPcVFWV5BHAl6pq23kKed4MUlftfKsB3wZuprnZwzFzH+38G3Db2gV4S1U9e36inJwt1wOqqpNp7lwyiBcBXxhiOAveAPVVwLpJAqzTzrtiLmJbaAaoq8XA99p5fwEsSnKfuYhtoamq31bVme3rG4DzaZ7o2msPmh8qVVU/ATZoD9arnEHqq6p+X1WnA7fOQ4gLxoB1dUpVXdMO/oTmOQ2rpAHr68ae2+jek+a4v8oZ8LgFTaPAscDv5zC8BWca9bVgmVx3LMk9gF1pdhBN7CPAQ4DLgXOBN1bV7fMb0oJ1NvDXAEl2pGllXGW/1MckWQQ8Cvhp36TNgEt6hi9lxA7MwzBJfanPgHX1auAbcxLQAjdZfSV5XpJfAMcDr5rj0BacieoqyWbA84CPzUNYC9YU++Jj2q6l30jy0LmNbHIm193bHfhxVQ3ayr2qegZwFrAp8EjgI0nWm8+AFrCDaVpfz6Jp2fgZcNu8RjTP2tPNxwJvqqrr5zuehc76GtwgdZXkSTTJ9dvmMraFaKr6qqqvtF1Bngu8Z47DW1CmqKv/BN5mI9OdpqivM4Gtqmo74L+Ar85xeJOa84fIrAL2ZhXvEjKgVwIHt6cMlyf5NbAtcNr8hrXwtAeVV0JzsR7wa+DCeQ1qHiVZg+aA+/mq+vI4s1wGbNEzvHk7bpU0QH2pNUhdtX2HPwnsVlVXzWV8C810tq2qOjnJ/ZNsXFV/mJsIF44B6moJcHRziGdj4JlJVlTVV+cuyoVjqvrqTbar6oQkH11I25Yt1x1Ksj7wROBr8x3LCPgN8BSAtv/wg1mFE8bJJNkgyZrt4GuAk1fV1sf2x8WnaC4qO2SC2Y4DXpbGXwHXVdVv5yzIBWTA+hKD1VWSLYEvAy+tql/OZXwLzYD19cB2PtLctWctYJX7QTJIXVXV1lW1qKoWAccAf7cKJ9aDbFv37dm2dqTJZxfMtmXL9YCSfAHYBdg4yaXAu4A1AKrq4+1szwO+VVU3zUuQC8gA9fUe4PAk5wKhOR22IH5xzrUB6uohwBFJClhGczp6VfU44KXAuW03GYB3AFvCHfV1As2dQpYDf6Rt9V9FTVlfSe4LLAXWA25P8iZg8Sr4A26QbesAYCPgo+33+oqqWjL3oS4Ig9TX82l+6N4K/AnYq+cCx1XJIHWlOw1SX3sCf5tkBc22tfdC2ra8FZ8kSZLUEbuFSJIkSR0xuZYkSZI6YnItSZIkdcTkWpIkSeqIybUkSZLUEZNrSZIkqSMm15IkSVJHTK4lSZKkjphcS5IkSR0xuZYkSZI6YnItSZIkdcTkWpIkSeqIybUkSZLUEZNrSZIkqSMm15IkSVJHTK4lSZKkjphcS5IkSR0xuZYkSZI6YnKtBSHJ4Un+ZRbL35jk/l3GNB+SLEpSSVaf71hmKskrkvxoSGW/I8knOyzvju1mttvgOGV/PMk/d1VeX9mLkyxNkmGUP8qSHJjkc+3rLdvPeLWp5l2VJLl7kv9Ncl2S/5nveGZrkOP/VNtCBzEsS7JLB+V8MMnfzj4izSeTa82pJCcluSbJWpPMs0uSS6dTblWtU1UXziCew5OsSHK/6S47H5JclORP7ZfEFW3868x3XF1ot42bk9yQ5PokZyTZv3dbqar3VtVrBixryvlmut2Ms76/+EFRVa+tqvfMtuwJvAf4QFVVkrWSfCrJxW3dnZVkt774npLkF0n+mOT7SbbqmbZWkk+3dX5FkjdPtuIkf9/Od3273Fo90xa15f+xXd9TB112GKrqN+1nfNtsy+r6x9c82xO4D7BRVb1gpoXMx4+T8fbtQfbjYW8LVfXQqjpptmUDHwDekWTNDsrSPDG51pxJsgh4PFDAc+ZwvRO1Wt0TeD5wHfCSuYqnA7tX1TrAI4FHAW+fr0CG0MK+X1WtC9wP+Adgb+CErltoR/zMwP2AJwFfbUetDlwCPBFYH3gn8KV2fyPJxsCXgX8G7gUsBb7YU+SBwDbAVm25/5hk1wnW/Qxgf+Ap7fz3B97dM8sXgJ8BGwH/BByTZJMBl71LG1aL6XS1cWwF/LKqVsx3PINK4y6fs1TVb4FfMIffkRqCqvLPvzn5Aw4AfgwcAny9b9rhwL8A9wT+BNwO3Nj+bQrsCJwKXAv8FvgIsGbP8gU8sKesjwEnADcBT50gnpfRJCVvBH7eN+1A4EvAZ4EbgGXAkp7pFwFvAc6hSc6/CKzdTnsF8KO+8nrjexZNAnJ9u/4De+Zb1M67+gQxX9T7foB/B47vGf4r4JS2ns4GdmnHPwk4t2e+bwOn9wz/EHhu+3p/4Fft+z4PeF7PfK9oP8P/AK5qP7ONgOPa93MaTavqj9r50877+3b6ucDDJnhvJwGv6Ru3JfBH4Nk9n8vn2tdrA59r47gWOJ2mNe5fgduAm2m2n4/0fAavA/4P+PUE283H27q5AfgBsNVEn8tYvMBD2nXd1q7v2t5tumf+/wcsB65u62vTvu3jtW1s1wKHAplku/3OFPvaOcDz29f7Aqf0TBvbx7Zthy8Hnt4z/T3A0ROUexTw3p7hpwBXtK8fBNwCrNu3Xb12qmXHWc83aH5o9Y47G/jr9vWHaPad64EzgMf37btj28hKnxuwdfu53tB+zh8Zm7ed/j/AFTT79MnAQ3vq8Fbgz+1n/L/t+Ie028G1NMeI5/Qd01Y6DgHPpNmnbgAuA94y4LHzcCbYNtvp27bTrgYuAF44SRw/bt/Hre17eXU736uA84FrgBP7yn9oT/m/A94B7NpXztmTHLPe3r7va4DPcOexckPg68CV7bSvA5v37WP/2sb8J+DzTLxvj+3Hdwc+CFzcfo4/asf1bwsnAf9Gc8y6HvgacK9ZbAsX0R6bgbWA/6TZty5vX6/VTtsFuJSm8eD3NN9nr+yrs38CPjPItuHfwvyb9wD8W3X+aBKLvwO2bw9O9+mZdjhtIjJ28OlbdnuaxHH19iB5PvCmnun9SdJ1wONozs6sPUE836VJTu8DrAC275l2YHsAfyawWnsQ/knP9Ivag/KmNK2B53NnEvEKJk+udwEe3sb2CJovq+e201b6Ahgn5t4D+OY0yeqH2uHNaBLNZ7ZlP60d3oTmy+VmYGNgjXadlwHrttP+RHOKGOAF7fu6G7AXzRfy/Xre2wrg9e1ncXfgaJofIvcEHtaWO5ZcP4Mm+dmAJtF+yFhZ47y3k+hLrtvxJwPv6/lcxhKnvwH+F7hH+xltD6w3UVltvX67/bzuPsF2cwPwBJovxw/1vI+/+Fx61zHBZ344d27TTwb+ADy6Lfu/gJP7Yvt6W09b0iQbu05QT+8HDp1kP7tP+1mPJc8fAj7WN8/Pac7abNiuu3df3JOeH2J9y50N7NUzvHG7/EbA84Dz++b/CPBfUy07znpeBvy4Z3gxTQI7lqC8pF3n6jRJyhXcmbD1biMrfW40P9APaT+DJ7Sfd29y/SqafWIsOTprvM+zHV6D5pj2DmDN9jO+AXjwRMchmkTq8e30DYFHD3jsPJyJt8170vzQeGVbH49qt7XFk8RxRx218+zRvpeHtGW8k/YHWVsfv23ree12eKf+up4k9ototrctaPa9H3PnfrERzXZ4j7bc/wG+2reP/YYmuV+9rfOTGH/fHtuPD23n2YzmuPDYts76t4WTaI5VD2vr8NiZbgvjHJsPAn4C3Jvm+HsK8J6e4/+Kdp41aI7XfwQ27Cnrr4EzB9k2/FuYf3f5UyxaGJLsTHMq8ktVdQZNy+iLB12+qs6oqp9U1Yqqugj4BM1p8Il8rap+XFW3V9XN48SzJU1r7lFV9TuaRPtlfbP9qKpOqKaP3pHAdn3TP1xVl1fV1TRJ3iMHfC8nVdW5bWzn0JxKn+y99PtqkhtovlB/D7yrHf8S4IQ25tur6ts0XQCeWVV/omnZfQJNEno2zZfc42h+tPxfVV3Vxvc/7fu6vaq+SNOaumPP+i+vqv+q5pTyn2m+HA+oqpuq6ufAET3z3krzBbUtTUvs+dWc9pyOy2m+lPvdSvPl/MCquq3dRq6foqx/q6qr2/oYz/FVdXJV3ULTevSYJFtMM97x7AN8uqrObMt+e1v2op55Dq6qa6vqN8D3mXh72oAm0foLSdagad07oqp+0Y5ehya56nUdzeeyTs9w/7Tx9Jc19nrdcab1lzXZsv2+Ajyyp2/4PsCX27qjqj5XVVe1x4MP0iRAD54gZuCOfX4H4J+r6paqOplmv71DVX26qm5o13MgsF2S9Sco8q/a93RwVf25qr5H8wPpRT3z9B+HbgUWJ1mvqq6pqjMni7nPRNvms4GLquozbX38jCZR7O1LPenxkOasyb+1++cK4L3cWf/PpjnD8MGqurmtn59OI25oWpgvaY+V/0pbR+1neGxV/bGqbmin9R8LD6+qZe17u3WylbTdRl4FvLGqLmuPC6eMbTfjOLKqfl5VN9F0m3rhWPedaW4L/fYBDqqq31fVlTTdn17aM/3WdvqtVXUCTQt47/Z7A81+rhFlcq258nLgW1X1h3b4qHbcQJI8KMnXxy6Gojn4bzzJIpdMUeRLaVrZzmqHPw+8uE1OxlzR8/qPwNp9fXX7pw90YWGSndqLvq5Mch3NF9tk76Xfc6vpl7wLTdI6tuxWwAuSXDv2B+xM038ZmlPJu9Ak2D+gabl5Yvv3g574XtZeFDdWxsP64uut2024s8/vmIvHXrQJx0doWpN+n+SwJOtN471C0wJ19Tjjj6Q5fX10ksuT/Hvf5zeeqbaLO6ZX1Y3tejedTrAT2JSV6+VGmrMKm/XMM+j2dA3jJKRtYnEkzQ+e/Xom3Qj01/l6NF/gN/YM908bT39ZY69vGGdaf1mTLbuSNtE6nqbPPTTJ2OfHpid5S5Lz09zt4lqavuZT7UObAte0idSYOz6TJKslOTjJr9pjzEXtpInK3RS4pKpu7yuv9zPt396eT9NSeXGSHyR5zBQx95po29wK2Klvv98HuO8kcfTbCvhQz/JX05xp2oymxflX04hz0thp6mhTgCT3SPKJ9mLc62nOUm3Q1z99qth7bUzTuj5ovP1xrQFsPINtod9K+zs977l1Va3c371/f1+X5kyNRpTJtYYuyd2BFwJPbJPjK4C/p2kJ6G8NhubUXb+P0VzksU1VrUdzKnayi9zGK6PXy4D798RzCM2B85lTLDeIm2hOcwKQ5L5904+i6XO7RVWtT9OXctoX7FXVD2hOT36gHXUJTUvMBj1/96yqg9vp/cn1D+hLrtuWqv+mSc42qqoNaE7p9sbXW7dX0pzi7G3d3bIvzg9X1fY0p/YfBLx10PfYtsxtT9N3t//931pV766qxTSnfp/NnWcfJvr8p9ou7ngfae7Cci+alvOxhOwePfP2fq5TlXs5TQIzVvY9aVrdL5tiufGcQ1OPd2gv+PwUTZeQ5/e18C2j56xLu+4HAMuq6hqaU/69++F27TLjWTbOvL9rz3oso9mn1u2bvmyAZcfzBeBFbQK6Nk1rPkkeD/wjzTFlw3YbvY6p96HfAhu2739M77b6YpruEU+lSdYXtePHyu3/jC8Htui7yG5LVv5MV1qmqk6vqj1ougt8laY71aAm2jYvAX7Qt9+vU1W9t3Obavu8BPibvjLuXlWntNMmus3dVOX+Rew0dXR5+/ofaFpsd2qP608Ye4uTrGOydf6BpkvUA2YY161tGdPdFvqttL+z8nsexENozi5qRJlcay48l+YilMU0p7ofSXPw+CF/2RUDmv7AG/WdgluX5qKTG5NsC8z4PqDtl/UDaLo6jMXzMJqkd7x4puts4KFJHplkrH9jr3WBq6vq5iQ7Mo3uMeP4T+Bp7Y+UzwG7J3lG2/KydprbGm7eznsKzRfZjsBpVbWMttWLpsUImr6HRZM0k+SVNHUzrmq6zHwZOLBthVpMzxmJJDu0LfVr0CSoN9NcrDqptqwn0lxkdBrNxVj98zwpycPbVq7rab4Yx8r+HRMnBJN5ZpKd09wG6z00/ewvaU/tXga8pK3bV7HyF/jvgM0z8e2zvgC8st0m1qI58/LTaro4Tde3gUe329aYj9HsU7uP0+XlK8DDkjy/XeYA4JyebiOfBd6ZZMN23/p/ND/aAEhz3/VdeuZ9dZr7bG9A0zf3cICq+iVwFvCudtt7Hs01BcdOtewETqDZPg8CvtjTQrwuzQ+6K4HVkxzAX7aY/4Wqupimm9S7k6yZpqva7j2zrEtzQeZVND+i3ttXRP829VOaFsd/TLJGW0e701yD8Bfade6TZP32x8/19OwLffU8nnG3TZquKA9K8tI2jjXa/e4hk9VHn48Db0/y0DaW9ZOMdSv5OnC/JG9Kc9vGdZPs1FMnizL1XTxel2TzJPei6dIydreadWmu97i2nfauiQroMeG+3W4jnwYOSbJpu68+JhPf8vEl7fZ4D5rt7Jj2mDbdbaHfF2j2qU3S3K3nAJrj86CeSHNRr0aUybXmwstprnz+TVVdMfZH011gn/TdFq390v8CcGF7mnJTmjtzvJjmFPJ/s/KtxGYSz9eq6ffcG8+HgGe3B/kZa5OMg4Dv0PRX7n+gyt8BB6XpN30A02u96l/XlTRJywHtF+0eNK36V9K0OL2Vdj9vT4efSdNi+ee2iFOBi6vq9+0859FcaX8qzRfIw2n6Zk9mP5pTmlfQJEuf6Zm2Hs3ndQ3NqdGraC7Im8hH2nr5Hc0Ph2NpLuwbLyG/L3AMTZJyPk3r+5HttA8Be6a5p/qHp4i/11E0X/BX07SY996i8f/R1OdVNBdYndIz7Xs0LbNXJPkDfarqOzR9Oo+laUF9AHd2eZiWaq4R+B7NZz12tuFvaH4kXpHmHug3Jtmnnf9Kmu4I/0rzOezUt+530ZxGv5imDt9fVd9sy96CZp87ty3rmzQXAX+f5kKzi1k5IdobWNKu52Bgz3b9gyzb/z5vofnh9lSaz2XMicA3gV+2ZdzM4F0HXty+/6vbdX+2Z9pn2/Iuo7mzxU/6lv0UTX/pa5N8td2Hdgd2o2nt/Cjwsp4fLeN5KXBR29XgtTTdN/6inicw7rbZdqF5Ok3dX06zH76Pph/6QKrqK+0yR7ex/bx9X2PlP619r1fQHNOe1C469gCaq5JM1n/8KOBbwIU029rYPaL/k+ai6D/Q1Pc3Bwh3qn37LTT1eDpNXb2PiXOdI2mOWVfQnB15Qzt+WtvCOOX+C80PuXPaWM7kzvc8qTS32lzMnbfa1AhK1aBndSRJC0F7huAIYMca4kE8yUtobkE2b/dSXxVMVc9JDqe5g9I75zSwDiS5iObuHt+Z71h6JTmJ5u4gnT3xtQtJPgj8qqo+Ot+xaOZG9kEKkrSqas8w7DAH61nlHg0+H6xnjamqf5jvGDR7dguRJEmSOmK3EEmSJKkjtlxLkiRJHbnL9LneeOONa9GiRfMdhiRJku7izjjjjD9U1SbjTbvLJNeLFi1i6dKl8x2GJEmS7uKSXDzRNLuFSJIkSR0xuZYkSZI6YnItSZIkdWSoyXWSXZNckGR5kv3Hmf6EJGcmWZFkz75pWyb5VpLzk5yXZNEwY5UkSZJma2jJdZLVgEOB3YDFwIvaR/b2+g3wCuCocYr4LPD+qnoIsCPw+2HFKkmSJHVhmHcL2RFYXlUXAiQ5GtgDOG9shqq6qJ12e++CbRK+elV9u53vxiHGKUmSJHVimN1CNgMu6Rm+tB03iAcB1yb5cpKfJXl/2xK+kiT7JlmaZOmVV17ZQciSJEnSzC3UCxpXBx4PvAXYAbg/TfeRlVTVYVW1pKqWbLLJuPfxliRJkubMMJPry4AteoY3b8cN4lLgrKq6sKpWAF8FHt1teJIkSVK3htnn+nRgmyRb0yTVewMvnsayGyTZpKquBJ4M+PhFSau0RfsfP9TyLzr4WUMtX5JWBUNruW5bnPcDTgTOB75UVcuSHJTkOQBJdkhyKfAC4BNJlrXL3kbTJeS7Sc4FAvz3sGKVJEmSujDMlmuq6gTghL5xB/S8Pp2mu8h4y34beMQw45MkSZK6tFAvaJQkSZJGjsm1JEmS1BGTa0mSJKkjJteSJElSR0yuJUmSpI6YXEuSJEkdMbmWJEmSOmJyLUmSJHXE5FqSJEnqiMm1JEmS1BGTa0mSJKkjJteSJElSR0yuJUmSpI6YXEuSJEkdMbmWJEmSOmJyLUmSJHXE5FqSJEnqiMm1JEmS1BGTa0mSJKkjJteSJElSR0yuJUmSpI4MNblOsmuSC5IsT7L/ONOfkOTMJCuS7DnO9PWSXJrkI8OMU5IkSerC0JLrJKsBhwK7AYuBFyVZ3Dfbb4BXAEdNUMx7gJOHFaMkSZLUpWG2XO8ILK+qC6vqz8DRwB69M1TVRVV1DnB7/8JJtgfuA3xriDFKkiRJnRlmcr0ZcEnP8KXtuCkluRvwQeAtU8y3b5KlSZZeeeWVMw5UkiRJ6sJCvaDx74ATqurSyWaqqsOqaklVLdlkk03mKDRJkiRpfKsPsezLgC16hjdvxw3iMcDjk/wdsA6wZpIbq+ovLoqUJEmSFophJtenA9sk2Zomqd4bePEgC1bVPmOvk7wCWGJiLUmSpIVuaN1CqmoFsB9wInA+8KWqWpbkoCTPAUiyQ5JLgRcAn0iybFjxSJIkScM2zJZrquoE4IS+cQf0vD6dprvIZGUcDhw+hPAkSZKkTi3UCxolSZKkkWNyLUmSJHXE5FqSJEnqiMm1JEmS1BGTa0mSJKkjJteSJElSR0yuJUmSpI6YXEuSJEkdMbmWJEmSOmJyLUmSJHXE5FqSJEnqiMm1JEmS1BGTa0mSJKkjJteSJElSR0yuJUmSpI6YXEuSJEkdMbmWJEmSOmJyLUmSJHXE5FqSJEnqiMm1JEmS1BGTa0mSJKkjQ02uk+ya5IIky5PsP870JyQ5M8mKJHv2jH9kklOTLEtyTpK9hhmnJEmS1IWhJddJVgMOBXYDFgMvSrK4b7bfAK8Ajuob/0fgZVX1UGBX4D+TbDCsWCVJkqQurD6dmZPcDVinqq4fYPYdgeVVdWG77NHAHsB5YzNU1UXttNt7F6yqX/a8vjzJ74FNgGunE68kSZI0l6ZsuU5yVJL1ktwT+DlwXpK3DlD2ZsAlPcOXtuOmJcmOwJrAr8aZtm+SpUmWXnnlldMtWpIkSerUIN1CFrct1c8FvgFsDbx0mEGNSXI/4EjglVV1e//0qjqsqpZU1ZJNNtlkLkKSJEmSJjRIcr1GkjVokuvjqupWoAZY7jJgi57hzdtxA0myHnA88E9V9ZNBl5MkSZLmyyDJ9ceBi4B7Aicn2QoYpM/16cA2SbZOsiawN3DcIEG1838F+GxVHTPIMpIkSdJ8mzS5bi9g/F1VbVZVz6yqornDx5OmKriqVgD7AScC5wNfqqplSQ5K8py2/B2SXAq8APhEkmXt4i8EngC8IslZ7d8jZ/geJUmSpDkx6d1Cqur2JP8IfKlnXAErBim8qk4ATugbd0DP69Npuov0L/c54HODrEOSJElaKAbpFvKdJG9JskWSe439DT0ySZIkacQMcp/rsacjvq5nXAH37z4cSZIkaXRNmVxX1dZzEYgkSZI06qZMrtvb8P0tzQWGACcBn2hvySdJkiSpNUi3kI8BawAfbYdf2o57zbCCkiRJkkbRIMn1DlW1Xc/w95KcPayAJEmSpFE1yN1CbkvygLGBJPcHbhteSJIkSdJoGqTl+q3A95NcCATYCnjVUKOSJEmSRtAgyfWPgG2AB7fDFwwvHEmSJGl0DdIt5NSquqWqzmn/bgFOHXZgkiRJ0qiZsOU6yX2BzYC7J3kUTZcQgPWAe8xBbJIkSdJImaxbyDOAVwCbA4f0jL8BeMcQY5IkSZJG0oTJdVUdARyR5PlVdewcxiRJkiSNpEEef35skmcBDwXW7hl/0DADkyRJkkbNlBc0Jvk4sBfwepp+1y+guR2fJEmSpB6D3C3ksVX1MuCaqno38BjgQcMNS5IkSRo9gyTXf2r//zHJpsCtwP2GF5IkSZI0mgZ5iMzXk2wAvB84Eyjgk8MMSpIkSRpFg1zQ+J725bFJvg6sXVXXDTcsSZIkafRM9hCZv55kGlX15eGEJEmSJI2myVqud2//3xt4LPC9dvhJwCmAybUkSZLUY8ILGqvqlVX1SmANYHFVPb+qnk9zv+s1Bik8ya5JLkiyPMn+40x/QpIzk6xIsmfftJcn+b/27+XTe1uSJEnS3BvkbiFbVNVve4Z/B2w51UJJVgMOBXYDFgMvSrK4b7bf0Dxi/ai+Ze8FvAvYCdgReFeSDQeIVZIkSZo3g9wt5LtJTgS+0A7vBXxngOV2BJZX1YUASY4G9gDOG5uhqi5qp93et+wzgG9X1dXt9G8Du/bEIEmSJC04g9wtZL/24sbHt6MOq6qvDFD2ZsAlPcOX0rRED2K8ZTfrnynJvsC+AFtuOWVjuiRJkjRUg7Rcj90ZZMFdwFhVhwGHASxZsqTmORxJkiSt4ibsc53kR+3/G5Jc3/N3Q5LrByj7MmCLnuHN23GDmM2ykiRJ0ryYsOW6qnZu/687w7JPB7ZJsjVNYrw38OIBlz0ReG/PRYxPB94+wzgkSaugRfsfP7SyLzr4WUMrW9Jom+whMveabMGxiw0nmb4iyX40ifJqwKeralmSg4ClVXVckh2ArwAbArsneXdVPbSqrk7yHpoEHeCgqdYnSZIkzbfJ+lyfARSQcaYVcP+pCq+qE4AT+sYd0PP6dJouH+Mt+2ng01OtQ5IkSVooJusWsvVcBiJJkiSNuoHuFtL2fd4GWHtsXFWdPKygJEmSpFE0ZXKd5DXAG2m6b5wF/BVwKvDkoUYmaZXgRWeSpLuSQR5//kZgB+DiqnoS8Cjg2mEGJUmSJI2iQZLrm6vqZoAka1XVL4AHDzcsSZIkafQM0uf60iQbAF8Fvp3kGuDiYQYlSZIkjaIpk+uqel778sAk3wfWB7451KgkSZKkETTIBY0fBo6uqlOq6gdzEJMkSZI0kgbpc30G8M4kv0rygSRLhh2UJEmSNIqmTK6r6oiqeibNHUMuAN6X5P+GHpkkSZI0YgZpuR7zQGBbYCvgF8MJR5IkSRpdUybXSf69bak+CDgXWFJVuw89MkmSJGnEDHIrvl8Bj6mqPww7GEmSJGmUDXIrvk/MRSCSJEnSqJtOn2tJkiRJk5gwuU6y9VwGIkmSJI26yVqujwFI8t05ikWSJEkaaZP1ub5bkncAD0ry5v6JVXXI8MKSJEmSRs9kLdd7A7fRJODrjvMnSZIkqceELddVNfY0xnOq6htzGJMkSZI0kga5W8gpSQ5JsrT9+2CS9YcemSRJkjRiBkmuPw3cALyw/bse+MwghSfZNckFSZYn2X+c6Wsl+WI7/adJFrXj10hyRJJzk5yf5O0DvyNJkiRpngzyhMYHVNXze4bfneSsqRZKshpwKPA04FLg9CTHVdV5PbO9Grimqh6YZG/gfcBewAuAtarq4UnuAZyX5AtVddFA70qSJEmaB4O0XP8pyc5jA0keB/xpgOV2BJZX1YVV9WfgaGCPvnn2AI5oXx8DPCVJgALumWR14O7An2lazCVJkqQFa5CW69cCn+3pZ30N8PIBltsMuKRn+FJgp4nmqaoVSa4DNqJJtPcAfgvcA/j7qrq6fwVJ9gX2Bdhyyy0HCEmSJEkanimT66o6G9guyXrt8Fy0IO9IcxvATYENgR8m+U5VXdgX22HAYQBLliypOYhLkiRJmtAg3UKAJqmeZmJ9GbBFz/Dm7bhx52m7gKwPXAW8GPhmVd1aVb8Hfgwsmca6JUmSpDk3SLeQmTod2CbJ1jRJ9N40SXOv42i6mJwK7Al8r6oqyW+AJwNHJrkn8FfAfw4xVkkaikX7Hz/fIUiS5tDALdfTVVUrgP2AE4HzgS9V1bIkByV5Tjvbp4CNkiwH3gyM3a7vUGCdJMtokvTPVNU5w4pVkiRJ6sJALddJHgss6p2/qj471XJVdQJwQt+4A3pe30xz273+5W4cb7wkSZK0kE2ZXCc5EngAcBbNRYbQ3CpvyuRakiRJWpUM0nK9BFhcVd6NQ5IkSZrEIMn1z4H70txzWpJ0F9XVxZcXHfysTsqRpFE0SHK9Mc3jx08DbhkbWVXPmXgRSZIkadUzSHJ94LCDkCRJku4KBnlC4w+S3AfYoR11WvtgF0mSJEk9BrlbyAuB9wMnAQH+K8lbq+qYIccmSfPCB79IkmZqkG4h/wTsMNZanWQT4DuAybUkSZLUY5AnNN6trxvIVQMuJ0mSJK1SBmm5/maSE4EvtMN70ffURUmSJEmDXdD41iTPBx7Xjjqsqr4y3LAkSZKk0TNIyzVVdSxw7JBjkSTdBXR5QagPpJE0aiZMrpP8qKp2TnID0Pvo8wBVVesNPTpJkiRphEyYXFfVzu3/decuHEmSJGl0TXnXjyRHDjJOkiRJWtUNcku9h/YOJFkd2H444UiSJEmja8LkOsnb2/7Wj0hyfft3A/A74GtzFqEkSZI0IiZMrqvq34D1gc9W1Xrt37pVtVFVvX3uQpQkSZJGw6S34quq25PsMFfBSFKXurwlnCRJgxikz/WZJtiSJEnS1AZJrncCTk3yqyTnJDk3yTmDFJ5k1yQXJFmeZP9xpq+V5Ivt9J8mWdQz7RFJTk2yrF3n2gO/K0mSJGkeDPKExmfMpOAkqwGHAk8DLgVOT3JcVZ3XM9urgWuq6oFJ9gbeB+zV3pHkc8BLq+rsJBsBt84kDkmSJGmuTJlcV9XFSbYDHt+O+mFVnT1A2TsCy6vqQoAkRwN7AL3J9R7Age3rY4CPJAnwdOCcsfVU1VUDrE+SdBdjv3lJo2aQh8i8Efg8cO/273NJXj9A2ZsBl/QMX9qOG3eeqloBXAdsBDwIqCQnJjkzyT9OENu+SZYmWXrllVcOEJIkSZI0PIN0C3k1sFNV3QSQ5H3AqcB/DTmunYEdgD8C301yRlV9t3emqjoMOAxgyZIlNcR4JEmSpCkNckFjgNt6hm9rx03lMmCLnuHN23HjztP2s14fuIqmlfvkqvpDVf0ROAF49ADrlCRJkubNIMn1Z4CfJjkwybuBnwCfGmC504FtkmydZE1gb+C4vnmOA17evt4T+F5VFXAi8PAk92iT7ieycl9tSZIkacEZ5ILGQ5KcRNNNo4BXVtXPBlhuRZL9aBLl1YBPV9WyJAcBS6vqOJok/cgky4GraRJwquqaJIfQJOgFnFBVXtUiSZKkBW2QPtdjQpPoDtIlBICqOoGmS0fvuAN6Xt8MvGCCZT9Hczs+SZIkaSQMcreQA4AjgA2BjYHPJHnnsAOTJEmSRs0gLdf7ANu1rcwkORg4C/iXIcYlSZIkjZxBLmi8HOh99Pha/OVdPyRJkqRV3iAt19cBy5J8m6bP9dOA05J8GKCq3jDE+CRJkqSRMUhy/ZX2b8xJwwlFkiRJGm2D3IrviPY+1Q9qR11QVbcONyxJkiRp9EyZXCfZheZuIRfR3IZviyQvr6qThxqZJEmSNGIG6RbyQeDpVXUBQJIHAV8Ath9mYJIkSdKoGeRuIWuMJdYAVfVLYI3hhSRJkiSNpkFars9I8knufFriPsDS4YUkSZIkjaZBkuvXAq8Dxm6590Pgo0OLSJIkSRpRkybXSVYDzq6qbYFD5iYkSZIkaTRN2ue6qm4DLkiy5RzFI0mSJI2sQbqFbEjzhMbTgJvGRlbVc4YWlSRJkjSCBkmu/3noUUiSJEl3ARMm10nWprmY8YHAucCnqmrFXAUmSZIkjZrJ+lwfASyhSax3o3mYjCRJkqQJTNYtZHFVPRwgyaeA0+YmJEmSJGk0TdZyfevYC7uDSJIkSVObrOV6uyTXt68D3L0dDlBVtd7Qo5MkSZJGyITJdVWtNpeBSJIkSaNu0ofIzFaSXZNckGR5kv3Hmb5Wki+203+aZFHf9C2T3JjkLcOMU5IkSerC0JLr9tHph9LcaWQx8KIki/tmezVwTVU9EPgP4H190w8BvjGsGCVJkqQuDbPlekdgeVVdWFV/Bo4G9uibZw+aW/4BHAM8JUkAkjwX+DWwbIgxSpIkSZ0ZZnK9GXBJz/Cl7bhx52nvSHIdsFGSdYC3Ae+ebAVJ9k2yNMnSK6+8srPAJUmSpJkYap/rWTgQ+I+qunGymarqsKpaUlVLNtlkk7mJTJIkSZrAZLfim63LgC16hjdvx403z6VJVgfWB64CdgL2TPLvwAbA7UlurqqPDDFeSZIkaVaGmVyfDmyTZGuaJHpv4MV98xwHvBw4FdgT+F5VFfD4sRmSHAjcaGItSZKkhW5oyXVVrUiyH3AisBrw6apaluQgYGlVHQd8CjgyyXLgapoEXJKkBW3R/sd3VtZFBz+rs7Ikzb9htlxTVScAJ/SNO6Dn9c3AC6Yo48ChBCdJkiR1bKFe0ChJkiSNHJNrSZIkqSMm15IkSVJHTK4lSZKkjphcS5IkSR0xuZYkSZI6MtRb8UmSpIWvy/t2j8d7eWtVYsu1JEmS1BGTa0mSJKkjdguRNG3DPoUsrUp8lLp012LLtSRJktQRk2tJkiSpIybXkiRJUkdMriVJkqSOmFxLkiRJHTG5liRJkjpici1JkiR1xORakiRJ6ogPkZEk6S7CBzxJ88+Wa0mSJKkjJteSJElSR4aaXCfZNckFSZYn2X+c6Wsl+WI7/adJFrXjn5bkjCTntv+fPMw4JUmSpC4MLblOshpwKLAbsBh4UZLFfbO9Grimqh4I/Afwvnb8H4Ddq+rhwMuBI4cVpyRJktSVYbZc7wgsr6oLq+rPwNHAHn3z7AEc0b4+BnhKklTVz6rq8nb8MuDuSdYaYqySJEnSrA3zbiGbAZf0DF8K7DTRPFW1Isl1wEY0Lddjng+cWVW39K8gyb7AvgBbbrlld5FLd0HeRUCSpOFb0Bc0JnkoTVeRvxlvelUdVlVLqmrJJptsMrfBSZIkSX2GmVxfBmzRM7x5O27ceZKsDqwPXNUObw58BXhZVf1qiHFKkiRJnRhmcn06sE2SrZOsCewNHNc3z3E0FywC7Al8r6oqyQbA8cD+VfXjIcYoSZIkdWZoyXVVrQD2A04Ezge+VFXLkhyU5DntbJ8CNkqyHHgzMHa7vv2ABwIHJDmr/bv3sGKVJEmSujDUx59X1QnACX3jDuh5fTPwgnGW+xfgX4YZmyRJktS1BX1BoyRJkjRKTK4lSZKkjphcS5IkSR0xuZYkSZI6YnItSZIkdcTkWpIkSerIUG/FJ2l2Fu1//HyHIEmSpsGWa0mSJKkjtlxLkqSh6uos3EUHP6uTcqRhsuVakiRJ6ogt11LH7CctSdKqy5ZrSZIkqSO2XEuSpJHQ5ZlB+29rWGy5liRJkjpiy7UkSdI0DPPaGlvUR58t15IkSVJHbLmWJEmrHO/spGGx5VqSJEnqiMm1JEmS1BGTa0mSJKkjJteSJElSR4Z6QWOSXYEPAasBn6yqg/umrwV8FtgeuArYq6ouaqe9HXg1cBvwhqo6cZixSpIkzTcflDP6hpZcJ1kNOBR4GnApcHqS46rqvJ7ZXg1cU1UPTLI38D5grySLgb2BhwKbAt9J8qCqum1Y8WrV5lXjkqS7moX63XZXT/qH2XK9I7C8qi4ESHI0sAfQm1zvARzYvj4G+EiStOOPrqpbgF8nWd6Wd+oQ49WIWagHDUmStOoaZnK9GXBJz/ClwE4TzVNVK5JcB2zUjv9J37Kb9a8gyb7Avu3gjUku6Cb0adsY+MM8rXvUWFfTY30NzrqaHutrcNbV9Fhfg1sl6yrvm/GiC6m+tppowkg/RKaqDgMOm+84kiytqiXzHccosK6mx/oanHU1PdbX4Kyr6bG+BmddTc+o1Ncw7xZyGbBFz/Dm7bhx50myOrA+zYWNgywrSZIkLSjDTK5PB7ZJsnWSNWkuUDyub57jgJe3r/cEvldV1Y7fO8laSbYGtgFOG2KskiRJ0qwNrVtI24d6P+BEmlvxfbqqliU5CFhaVccBnwKObC9YvJomAaed70s0Fz+uAF63wO8UMu9dU0aIdTU91tfgrKvpsb4GZ11Nj/U1OOtqekaivtI0FEuSJEmaLZ/QKEmSJHXE5FqSJEnqiMn1gJJ8Osnvk/x8gulvTXJW+/fzJLcluddcx7lQDFBf6yf53yRnJ1mW5JVzHeNCMUBdbZjkK0nOSXJakofNdYwLRZItknw/yXntdvPGceZJkg8nWd7W2aPnI9aFYMD62jbJqUluSfKW+YhzIRiwrvZpt6lzk5ySZLv5iHUhGLC+9mjr66wkS5PsPB+xzrdB6qpn3h2SrEiy51zGuJAMuG3tkuS6nrzrgPmIdSL2uR5QkicANwKfrapJk5skuwN/X1VPnpPgFqCp6ivJO4D1q+ptSTYBLgDuW1V/nuNQ590AdfV+4MaqeneSbYFDq+opcx3nQpDkfsD9qurMJOsCZwDPrarzeuZ5JvB64Jk0D676UFX1P8BqlTBgfd2b5mEIzwWuqaoPzEuw82zAunoscH5VXZNkN+BAt61J62sd4KaqqiSPAL5UVdvOU8jzZpC6audbDfg2cDPNTSCOmfto59+A29YuwFuq6tnzE+XkbLkeUFWdTHNHk0G8CPjCEMNZ8AaorwLWTRJgnXbeFXMR20IzQF0tBr7XzvsLYFGS+8xFbAtNVf22qs5sX98AnM9fPr11D5ofKlVVPwE2aA/Wq5xB6quqfl9VpwO3zkOIC8aAdXVKVV3TDv6E5hkMq6QB6+vGurMF7540x/1VzoDHLWgaBY4Ffj+H4S0406ivBcvkumNJ7gHsSrODaGIfAR4CXA6cC7yxqm6f35AWrLOBvwZIsiNNK+Mq+6U+Jski4FHAT/smbQZc0jN8KSN2YB6GSepLfQasq1cD35iTgBa4yeoryfOS/AI4HnjVHIe24ExUV0k2A54HfGwewlqwptgXH9N2Lf1GkofObWSTM7nu3u7Aj6tq0FbuVdUzgLOATYFHAh9Jst58BrSAHUzT+noWTcvGz4CFfN/3oWtPNx8LvKmqrp/veBY662twg9RVkifRJNdvm8vYFqKp6quqvtJ2BXku8J45Dm9BmaKu/hN4m41Md5qivs4Etqqq7YD/Ar46x+FNamgPkVmF7c0q3iVkQK8EDm5PGS5P8mtgW3wS519oDyqvhOZiPeDXwIXzGtQ8SrIGzQH381X15XFmuQzYomd483bcKmmA+lJrkLpq+w5/Etitqq6ay/gWmulsW1V1cpL7J9m4qv4wNxEuHAPU1RLg6OYQz8bAM5OsqKqvzl2UC8dU9dWbbFfVCUk+upC2LVuuO5RkfeCJwNfmO5YR8BvgKQBt/+EHswonjJNJskGSNdvB1wAnr6qtj+2Pi0/RXFR2yASzHQe8LI2/Aq6rqt/OWZALyID1JQarqyRbAl8GXlpVv5zL+BaaAevrge18pLlrz1rAKveDZJC6qqqtq2pRVS0CjgH+bhVOrAfZtu7bs23tSJPPLphty5brASX5ArALsHGSS4F3AWsAVNXH29meB3yrqm6alyAXkAHq6z3A4UnOBUJzOmxB/OKcawPU1UOAI5IUsIzmdPSq6nHAS4Fz224yAO8AtoQ76usEmjuFLAf+SNvqv4qasr6S3BdYCqwH3J7kTcDiVfAH3CDb1gHARsBH2+/1FVW1ZO5DXRAGqa/n0/zQvRX4E7BXzwWOq5JB6kp3GqS+9gT+NskKmm1r74W0bXkrPkmSJKkjdguRJEmSOmJyLUmSJHXE5FqSJEnqiMm1JEmS1BGTa0mSJKkjJteSNENJnpukkmw7D+u+KMnGc73eySRZlOTnE0y7X5Kvt693SXJdkrN6/vbqeX1Fkst6htdMcmNfea9I8pH29X5JVvlHa0taGLzPtSTN3IuAH7X/3zXPsfyFJKtV1W1DLH/1qlox4OxvBv67Z/iHVfXsvnm+2JZ7IHBjVX2gZ12Tlf1p4Mftf0maV7ZcS9IMJFkH2JnmoT5794zfJclJSY5J8oskn+95kthFSd6d5Mwk5461eCc5MMlbesr4eZJF7euvJjkjybIk+w4Q141JPpjkbOAxSV6S5LS2BfgTSVZL8oIkh7TzvzHJhe3r+yf5cfv6gCSnt7Ec1vMeTkryn0mWAm9Msn2Ss9v1vW6S0J4PfHPgCp6GqvojcFH7pDZJmlcm15I0M3sA32wfg31Vku17pj0KeBOwGLg/zRPHxvyhqh4NfAx4C1N7VVVtDywB3pBkoynmvyfw06rajuZxwHsBj6uqRwK3AfsAPwQe387/+Db+zdrXJ7fjP1JVO1TVw4C7A72tzGtW1ZKq+iDwGeD17frGlWRr4JqquqVn9OP7uoU8YIr3dffe+YGD+qYv7XlPkjRvTK4laWZeBBzdvj66HR5zWlVdWlW3A2cBi3qmfbn9f0bf+Im8oW0V/gmwBbDNFPPfBhzbvn4KsD1wepuQPgW4f1VdAayTZN22zKOAJ9Akpz9sl31Skp8mORd4MvDQnnWMdd/YANigqsYS8iMniOl+wJV9435YVY/s+fvVFO/rT73z0zyKvNfvgU2nKEOShs4+15I0TUnuRZNwPjxJAasBleSt7Sy9LbS3sfKx9pZxxq9g5caOtdv17AI8FXhMVf0xyUlj0yZxc08/6wBHVNXbx5nvFOCVwAU0CfWrgMcA/5BkbeCjwJKquqTtA9273pumiKHfnwaIe7bWbtcjSfPKlmtJmr49gSOraquqWlRVWwC/ZubdEi4CHg2Q5NHA1u349Wm6U/yx7Z/9V9Ms97vAnknu3ZZ9ryRbtdN+SNMt5WTgZ8CTgFuq6jruTIT/0PYt33O8wqvqWuDaJDu3o/aZII5fMlgr/Ww8CBj3TiWSNJdMriVp+l4EfKVv3LGs3DVkOo4F7pVkGbAfTTIKzQWAqyc5HziYpmvIwKrqPOCdwLeSnAN8m6aLBjTJ9RbAyW1L9yU0dz4ZS5r/myZZPRE4fZLVvBI4tO12Mu4tParqJuBXSR7YM7q/z/W4Cfw0PI7m/UnSvEpVzXcMkqS7uCTPA7avqncOoexHAW+uqpd2XbYkTZd9riVJQ1dVXxngTicztTHwz0MqW5KmxZZrSZIkqSP2uZYkSZI6YnItSZIkdcTkWpIkSeqIybUkSZLUEZNrSZIkqSP/H7HL4/eOAH3KAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 864x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# model annual rewards for perfect participation\n",
    "\n",
    "import math\n",
    "import pandas as pd\n",
    "\n",
    "def get_quantile(pmf, quantile):\n",
    "    cumulative = 0\n",
    "    for x, prob in sorted(pmf.items()):\n",
    "        cumulative += prob\n",
    "        if cumulative >= quantile:\n",
    "            return x\n",
    "\n",
    "slots_per_year = SECONDS_PER_YEAR / SECONDS_PER_SLOT\n",
    "\n",
    "GWEI_PER_ETH = int(1e9)\n",
    "gwei_per_validator = 32 * GWEI_PER_ETH\n",
    "BASE_REWARD_FACTOR = 64\n",
    "PREALTAIR_NUM_BASE_REWARDS = 4\n",
    "PREALTAIR_PROPOSER_REWARD_QUOTIENT = 8\n",
    "\n",
    "base_reward = gwei_per_validator * BASE_REWARD_FACTOR // math.isqrt(NUM_VALIDATORS * gwei_per_validator)\n",
    "total_reward = base_reward * NUM_VALIDATORS\n",
    "\n",
    "prior_proposer_share = base_reward // PREALTAIR_NUM_BASE_REWARDS // PREALTAIR_PROPOSER_REWARD_QUOTIENT\n",
    "prior_proposer_reward = prior_proposer_share * NUM_VALIDATORS // SLOTS_PER_EPOCH\n",
    "prior_att_reward = base_reward - prior_proposer_share\n",
    "\n",
    "altair_proposer_reward = total_reward * PROPOSER_WEIGHT // SLOTS_PER_EPOCH // WEIGHT_DENOMINATOR\n",
    "altair_att_reward = base_reward * (HEAD_WEIGHT + SOURCE_WEIGHT + TARGET_WEIGHT) // WEIGHT_DENOMINATOR\n",
    "sync_reward = total_reward * COMMITTEE_EPOCHS * SYNC_WEIGHT // COMMITTEE_VALIDATORS // WEIGHT_DENOMINATOR\n",
    "\n",
    "# distribution of committee selections per year\n",
    "n_committees = [el for el in range(11)]\n",
    "pmf_committees = binom.pmf(n_committees, committees_per_year, COMMITTEE_VALIDATORS / NUM_VALIDATORS)\n",
    "\n",
    "# distribution of block proposal opportunities per year\n",
    "n_proposals = [el for el in range(51)]\n",
    "pmf_proposals = binom.pmf(n_proposals, slots_per_year, 1 / NUM_VALIDATORS)\n",
    "\n",
    "n_bins = 32\n",
    "bins = [1.7 + i / 40 for i in range(n_bins)]\n",
    "altair_hist = [0] * n_bins\n",
    "prior_hist = [0] * n_bins\n",
    "\n",
    "# calculate all possible reward levels (up to 50 block proposals) assuming perfect participation\n",
    "prior_pmf = {}\n",
    "for props in n_proposals:\n",
    "    reward = props * prior_proposer_reward + epochs_per_year * prior_att_reward\n",
    "    prior_pmf[reward] = pmf_proposals[props]\n",
    "\n",
    "# bin the rewards to generate histogram\n",
    "for reward_gwei, prob in prior_pmf.items():\n",
    "    reward = reward_gwei / GWEI_PER_ETH\n",
    "    for i, edge in enumerate(bins[1:]):\n",
    "        if reward < edge:\n",
    "            prior_hist[i] += prob\n",
    "            break\n",
    "\n",
    "prior_mean = sum([p * r / GWEI_PER_ETH for r, p in prior_pmf.items()])\n",
    "prior_sigma = math.sqrt(sum([p * (r / GWEI_PER_ETH)**2 for r, p in prior_pmf.items()]) - prior_mean**2)\n",
    "prior_lq = get_quantile(prior_pmf, 0.25) / GWEI_PER_ETH\n",
    "prior_median = get_quantile(prior_pmf, 0.5) / GWEI_PER_ETH\n",
    "prior_uq = get_quantile(prior_pmf, 0.75) / GWEI_PER_ETH\n",
    "prior_iqr = prior_uq - prior_lq\n",
    "\n",
    "print('Pre-altair annual reward statistics (ETH)')\n",
    "print('-----------------------------------------')\n",
    "print(f'             median: {prior_median:.4f}')\n",
    "print(f'               mean: {prior_mean:.4f}')\n",
    "print(f' standard deviation: {prior_sigma:.4f}')\n",
    "print(f'interquartile range: {prior_iqr:.4f}')\n",
    "#print(sum(prior_hist)) # check histogram sums to unity\n",
    "\n",
    "# calculate all possible reward levels (up to 50 block proposals and 10 committee selections)\n",
    "altair_pmf = {}\n",
    "for comms in n_committees:\n",
    "    for props in n_proposals:\n",
    "        reward = comms * sync_reward + props * altair_proposer_reward + epochs_per_year * altair_att_reward\n",
    "        prob = pmf_committees[comms] * pmf_proposals[props]\n",
    "        if reward in altair_pmf:\n",
    "            altair_pmf[reward] += prob\n",
    "        else:\n",
    "            altair_pmf[reward] = prob\n",
    "\n",
    "# bin the rewards to generate histogram\n",
    "for reward_gwei, prob in altair_pmf.items():\n",
    "    reward = reward_gwei / GWEI_PER_ETH\n",
    "    for i, edge in enumerate(bins[1:]):\n",
    "        if reward < edge:\n",
    "            altair_hist[i] += prob\n",
    "            break\n",
    "\n",
    "altair_mean = sum([p * r / GWEI_PER_ETH for r, p in altair_pmf.items()])\n",
    "altair_sigma = math.sqrt(sum([p * (r / GWEI_PER_ETH)**2 for r, p in altair_pmf.items()]) - altair_mean**2)\n",
    "altair_lq = get_quantile(altair_pmf, 0.25) / GWEI_PER_ETH\n",
    "altair_median = get_quantile(altair_pmf, 0.5) / GWEI_PER_ETH\n",
    "altair_uq = get_quantile(altair_pmf, 0.75) / GWEI_PER_ETH\n",
    "altair_iqr = altair_uq - altair_lq\n",
    "\n",
    "print('\\nAltair annual reward statistics (ETH)')\n",
    "print('-------------------------------------')\n",
    "print(f'             median: {altair_median:.4f}')\n",
    "print(f'               mean: {altair_mean:.4f}')\n",
    "print(f' standard deviation: {altair_sigma:.4f}')\n",
    "print(f'interquartile range: {altair_iqr:.4f}')\n",
    "#print(sum(altair_hist)) # check histogram sums to unity\n",
    "\n",
    "print(f'\\nrelative spread: {altair_sigma / prior_sigma:.1f} (standard deviation) / '\n",
    "      f'{altair_iqr / prior_iqr:.1f} (interquartile range)')\n",
    "\n",
    "fig, (ax1,ax2) = plt.subplots(2, 1, figsize=(12,10))\n",
    "ax1.bar(bins, prior_hist, 1 / n_bins, align='edge')\n",
    "ax1.set_title('Pre-Altair Annual Rewards Distribution (200,000 validators, perfect participation)')\n",
    "ax1.set_ylabel('Proportion of validators')\n",
    "ax2.bar(bins, altair_hist, 1 / n_bins, align='edge')\n",
    "ax2.set_title('Altair Annual Rewards Distribution (200,000 validators, perfect participation)')\n",
    "ax2.set_xlabel('Annual reward (ETH)')\n",
    "ax2.set_ylabel('Proportion of validators')\n",
    "\n",
    "ax.set_title('Distribution of annual rewards assuming perfect performance')\n",
    "ax.set_xlabel('Annual reward (ETH)')\n",
    "ax.set_ylabel('Proportion of validators');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As expected, the statistics and charts above show that with perfect participation, although the mean issuance is unchanged by Altair, the spread of rewards is increased significantly, with a 5.7 factor increase in the standard deviation of rewards in Altair, as compared with the pre-Altair spread.\n",
    "\n",
    "# Penalties\n",
    "\n",
    "The changes in reward weights in Altair are matched by equivalent changes to the weights of penalties for missed or incorrect attestations. But more significant than the weightings is the way in which delayed attestations are treated. If an attestation is not included in the earliest possible slot, then the validator will be treated as not having participated for some or all of the components of the attestation, leading to penalties for those components that were late.\n",
    "\n",
    "This is an important difference. Consider for example an attestation which is correct, but is included one slot 'late'. Before Altair, this attestation would receive the maximum reward for the source, target and head votes, and half of the 'delay' reward. Altogether this would have amounted to approximately 90% of the maximum available reward for the attesation. However, under Altair's rules, the validator will be treated as though the head vote was incorrect, and will be penalised. Therefore, at best an attestation which is one slot late will receive 48% of the maximum available reward.\n",
    "\n",
    "Attestations which are too late to receive the source reward (i.e. with a delay of greater than 5 slots) will result at best in zero total reward under Altair (and may be overall negative), whereas the same attestation pre-Altair would have received up to 77% of the maximum reward. In short, penalties for late attestations are much more severe under Altair. Under situations of network stress as have frequently been observed on testnets with lower-participation rates, even completely reliable validators are likely to be punished.\n",
    "\n",
    "## Slashing and Inactivity Penalties\n",
    "\n",
    "In addition to the penalty weights and delay mechanism a few changes have been made to the slashing and inactivity leak parameters. Since neither of these factors are being modelled here (the conditions for inactivity leak have never occurred on mainnet, and slashing is rare and easily avoided by users with a standard set up), they will not be covered here. See [Vitalik Buterin's annotated spec](https://github.com/ethereum/annotated-spec/blob/master/altair/beacon-chain.md#introduction) for further details.\n",
    "\n",
    "# Mainnet Data\n",
    "\n",
    "For a more realistic direct comparison between the pre-Altair and post-Altair reward schemes, we'd like to have some real world data. Fortunately, the existing mainnet data is similar enough that, if we make a few assumptions, we can compare how validators would have been rewarded under Altair, with the rewards they actually received.\n",
    "\n",
    "## Methodology\n",
    "\n",
    "For ease, we're going to use the data from the first 62,000 epochs of the beacon chain (the same dataset as in the previous article). This means our data covers the first 8 months of the beacon chain's operation up until [3 September](https://beaconcha.in/epoch/62000). Then, for each epoch, we'll calculate the rewards available for correct (and timely) source, target and head votes. Using the epoch summary data provided by [chaind](https://github.com/wealdtech/chaind), we can see whether each vaildator voted correctly, and how many slots delayed the attestation was included. We can use this information, along with the updated `base_reward` calculation and weightings, to work out what rewards and/or penalties the validator would have received, had the Altair reward scheme been in place.\n",
    "\n",
    "Second, we need to simulate the sync committee and make some guess as to how each vaildator would have performed, if selected for a sync committee. For this analysis, 512 validators are selected at random every 256 epochs, and their performance in the sync committee for each epoch is assumed to perfect if they successfully attested in that epoch, or non-existent otherwise.\n",
    "\n",
    "Finally, we can use the previously-calculated proposer rewards. These are simply multiplied by 4 to get the proposer reward which would have applied under Altair. These steps have been implemented in a [Python script](altair_sim.py) with the results saved to JSON files.\n",
    "\n",
    "## Assumptions\n",
    "\n",
    "So, clearly a number of assumptions are made in order to do this comparison, such as:\n",
    "\n",
    "* the blocks proposed are identical, containing the same number of attestations and therefore worth exactly 4 times as much under Altair compared with the previous rules\n",
    "* attestation inclusion occurs at the same time under Altair as it did under the previous rules (i.e. changes to epoch processing, client optimisations etc. which may have altered the speed at which attestations are included are not considered here)\n",
    "\n",
    "And the by far the strongest assumption:\n",
    "\n",
    "* each sync committee participant would have successfully performed their duties for the entire epoch if in that epoch the validator submitted an attestation\n",
    "\n",
    "This last assumption is the shakiest one — we're assuming that validators who successfully submitted an attestation, even if delayed, would have successfully participated for the whole epoch (i.e. 32 times). In some cases, then, this will be a generous assumption about validators' performance (even if they submitted an attestation that epoch, we cannot be sure that they would have flawlessly participated in the sync committee for each slot). On the other hand, in some cases this will be an excessively harsh assumption, since validators who failed to attest in a given epoch might still have successfully participated in the sync committee for some or all of the slots.\n",
    "\n",
    "With all this in mind, let's take a look at the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "code_folding": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Statistics — first 62,000 epochs (pre-Altair rewards scheme, measured in ETH)\n",
      "-----------------------------------------------------------------------------\n",
      "             median: 2.1649\n",
      "               mean: 2.1387\n",
      " standard deviation: 0.1459\n",
      "interquartile range: 0.0551\n",
      "\n",
      "Statistics — first 62,000 epochs (Altair rewards scheme, measured in ETH)\n",
      "-------------------------------------------------------------------------\n",
      "             median: 2.1241\n",
      "               mean: 2.1144\n",
      " standard deviation: 0.1775\n",
      "interquartile range: 0.1344\n",
      "\n",
      "Comaprison\n",
      "-----------\n",
      "The mean per-validator reward under Altair changed by -1.1%\n",
      "The interquartile range (spread) of rewards under Altair was 2.4 times greater\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# calculate stats and plot data for net rewards under pre-Altair and Altair reward schemes\n",
    "\n",
    "import json\n",
    "\n",
    "with open('aggregate_rewards.json') as f:\n",
    "    rewards = json.load(f)\n",
    "with open('check.json') as f:\n",
    "    check = json.load(f)\n",
    "\n",
    "# we're using the \"reduced genesis set\" of validators as in the previous article\n",
    "\n",
    "with open('reduced_genesis_set.json') as f:\n",
    "    reduced_genesis_set = json.load(f)\n",
    "\n",
    "altair_rewards = []\n",
    "prior_rewards = []\n",
    "for validator_index in reduced_genesis_set:\n",
    "    altair_rewards += [rewards[str(validator_index)] / 1e9]\n",
    "    prior_rewards += [check[str(validator_index)] / 1e9]\n",
    "\n",
    "df = pd.DataFrame({'altair_rewards': altair_rewards, 'prior_rewards': prior_rewards})\n",
    "\n",
    "prior_iqr = df['prior_rewards'].quantile(0.75) - df['prior_rewards'].quantile(0.25)\n",
    "altair_iqr = df['altair_rewards'].quantile(0.75) - df['altair_rewards'].quantile(0.25)\n",
    "prior_mean = df[\"prior_rewards\"].mean()\n",
    "altair_mean = df[\"altair_rewards\"].mean()\n",
    "mean_diff = prior_mean - altair_mean\n",
    "\n",
    "print('Statistics — first 62,000 epochs (pre-Altair rewards scheme, measured in ETH)')\n",
    "print('-----------------------------------------------------------------------------')\n",
    "print(f'             median: {df[\"prior_rewards\"].quantile(0.5):.4f}')\n",
    "print(f'               mean: {prior_mean:.4f}')\n",
    "print(f' standard deviation: {df[\"prior_rewards\"].std():.4f}')\n",
    "print(f'interquartile range: {prior_iqr:.4f}')\n",
    "\n",
    "print('\\nStatistics — first 62,000 epochs (Altair rewards scheme, measured in ETH)')\n",
    "print('-------------------------------------------------------------------------')\n",
    "print(f'             median: {df[\"altair_rewards\"].quantile(0.5):.4f}')\n",
    "print(f'               mean: {altair_mean:.4f}')\n",
    "print(f' standard deviation: {df[\"altair_rewards\"].std():.4f}')\n",
    "print(f'interquartile range: {altair_iqr:.4f}')\n",
    "\n",
    "print('\\nComaprison')\n",
    "print('-----------')\n",
    "print(f'The mean per-validator reward under Altair changed by {100*(altair_mean / prior_mean - 1):.1f}%')\n",
    "print(f'The interquartile range (spread) of rewards under Altair was {altair_iqr / prior_iqr:.1f} times greater')\n",
    "\n",
    "fig, (ax1,ax2) = plt.subplots(2, 1, figsize=(12,10))\n",
    "bins = [b/20 for b in range(50)]\n",
    "df['prior_rewards'].plot.hist(ax=ax1, bins=bins)\n",
    "df['altair_rewards'].plot.hist(ax=ax2, bins=bins)\n",
    "ax1.set_title(\"Net rewards — pre-Altair reward scheme (first 62,000 epochs)\")\n",
    "ax1.set_ylabel(\"Number of validators\")\n",
    "ax2.set_title(\"Net rewards — Altair reward scheme (first 62,000 epochs)\")\n",
    "ax2.set_xlabel(\"Net reward (ETH)\")\n",
    "ax2.set_ylabel(\"Number of validators\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Comparing Validator Distributions\n",
    "\n",
    "Comparing the statisics, the Altair rules are less forgiving than the previous scheme, since the mean reward drops by 1.1% when they are applied to this dataset. As hinted earlier, this is most likely due to the impact of the harsher penalties for delayed attestations.\n",
    "\n",
    "Also as predicted, the spread of rewards is greater under Altair. Curiously, although clearly visible in the histograms above, this effect is not obvious in the comparison between standard deviations, which only increases slightly for Altair. This presumably is due to the influence of outliers (e.g. a few harshly penalised validators who have never submitted an attestation). However, when comparing the the more robust interquartile range, we see a spread over twice as large for rewards under Altair than the previous scheme. This is not as great as the 7 times increase in spread predicted from our modelling of perfect validators. This is because the variability in rewards in real data is due not only to the random allocation of proposer/sync duties, but also due to imperfect performance of validators, with the result that the real data in the pre-Altair case had a greater spread to start with than in the modelled perfect data, so the increase in variability from Altair is not so pronounced.\n",
    "\n",
    "## Comparing Issuance\n",
    "\n",
    "One possibility we should remember in using data which spans almost the entire period of the beacon chain's existence, is that network performance has varied over time. To see how the switch to Altair would have impacted on rewards, had it been active at different times in the beacon chain's history, the total per-epoch issuance has been calculated under both Altair and the pre-Altair rule. The relative change in issuance is plotted below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "code_folding": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean change in issuance: -1.05%\n",
      "Greatest per-epoch drop in issuance from Altair: -82.47%, greatest increase: 1.38%\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot relative issuance by epoch\n",
    "\n",
    "with open('issuance.json') as f:\n",
    "    issuance = json.load(f)\n",
    "\n",
    "delta = []\n",
    "sum_old = sum_new = 0\n",
    "for (old, new) in zip(issuance['old_issuance'], issuance['alt_issuance']):\n",
    "    delta.append(100 * (new / old - 1))\n",
    "\n",
    "print(f\"Mean change in issuance: {sum(delta) / len(delta):.2f}%\")\n",
    "print(f\"Greatest per-epoch drop in issuance from Altair: {min(delta):.2f}%, greatest increase: {max(delta):.2f}%\")\n",
    "    \n",
    "fig, ax = plt.subplots(figsize=(12, 8))\n",
    "ax.plot(pd.Series(delta).rolling(16).mean())\n",
    "ax.set_title('Percentage Change in Issuance after Applying Altair Rewards Scheme (16-epoch moving average)')\n",
    "ax.set_xlabel('Epoch')\n",
    "ax.set_ylabel('% change');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As shown above, the rewards under Altair are on average around 1% down on the status quo. However, the rewards dropped considerably more during the [April 2021 missing blocks incident](https://medium.com/prysmatic-labs/eth2-mainnet-incident-retrospective-f0338814340c). During this incident the dominant beacon chain client stopped producing blocks, resulting in delayed attestations across the network and the participation rate dropping to 84.8%. While the impact on rewards for the pre-Altair network was minimal, the impact of such an event is clearly far greater if Altair rules are applied. To a lesser extent, the same effect can be observed around genesis (when participation rates were also slightly reduced), and around epoch 59000, corresponding to the [August 2021 orphaned blocks incident](https://blog.lido.fi/lido-20-08-2021-orphaned-blocks-in-ethereum-incident-postmortem/).\n",
    "\n",
    "The reason such incidents have a greater impact under Altair than the status quo, is that delayed attestations are much more harshly punished under Altair, as explained previously. This is illustrated in the plot below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "code_folding": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot attestation reward against inclusion delay\n",
    "\n",
    "pre_altair_reward = []\n",
    "altair_reward = []\n",
    "x = []\n",
    "for delay in range(1,33):\n",
    "    x += [delay]\n",
    "    pre_altair_reward.append(0.75 + 0.25 * (7/8) / delay)\n",
    "    if delay == 1:\n",
    "        altair_reward.append((HEAD_WEIGHT + SOURCE_WEIGHT + TARGET_WEIGHT) / WEIGHT_DENOMINATOR)\n",
    "    elif delay <= 5:\n",
    "        altair_reward.append((SOURCE_WEIGHT + TARGET_WEIGHT - HEAD_WEIGHT) / WEIGHT_DENOMINATOR)\n",
    "    else:\n",
    "        altair_reward.append((TARGET_WEIGHT - HEAD_WEIGHT - SOURCE_WEIGHT) / WEIGHT_DENOMINATOR)\n",
    "        \n",
    "fig, ax = plt.subplots(figsize=(12, 8))\n",
    "ax.plot(x, pre_altair_reward, label=\"Pre-Altair\")\n",
    "ax.plot(x, altair_reward, label=\"Altair\")\n",
    "ax.set_xlabel(\"Inclusion Delay\")\n",
    "ax.set_ylabel(\"Per-Epoch Reward (as fraction of max issuance)\")\n",
    "ax.set_title(\"Reward for Correct Attestation According to Inclusion Delay\")\n",
    "ax.set_xlim([1,32])\n",
    "leg = ax.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Conclusions\n",
    "\n",
    "As was seen from our initial modelling, the four-fold increase in the proposer reward, and the introduction of the sync committee will  contribute to a greater variability of rewards than is currently the case. This variability will have a greater impact on solo validators than large staking pools (whose rewards will be closer to the average). This effect should be borne in mind when considering any future changes which could impact the structure of validator rewards (such as the future introduction of sharding).\n",
    "\n",
    "Additionally, on the basis of the analysis above, it appears likely that some reduction in mean rewards will occur when Altair goes live, even though the changes introduced are neutral in terms of the maximum rewards available. This impact will be felt most keenly by those validators whose attestations have a tendency to be included late, perhaps due to network latency. Validators would therefore be well advised to keep a close eye on their performance before and after Altair goes live.\n",
    "\n",
    "In particular, the rewards for validators are likely to drop significantly more under Altair during times of reduced participation or a failure to produce blocks, resulting in delayed inclusion of attestations. Such conditions have only occured once so far on the beacon chain, but we should expect that future incidents may occur as future updates (the Merge and sharding in particular) introduce new complexities.\n",
    "\n",
    "\n",
    "# Acknowledgements\n",
    "Many thanks to Lido finance for funding this work through their [ecosystem grants programme](https://research.lido.fi/t/legos-first-three-months/786). As ever, thanks also to Jim McDonald for the chaind data which made this analysis possible, and for valuable feedback from Barnabé Monnot and Vasiliy Shapovalov."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}