1 00:00:00,750 --> 00:00:03,050 Now we're going to talk about mining pools. 2 00:00:03,050 --> 00:00:05,090 So what is a Mining pool? 3 00:00:05,090 --> 00:00:09,130 Well let's look at the economics of being a small miner. 4 00:00:09,130 --> 00:00:14,150 So say you're an individual who spent $6,000 of your hard earned money 5 00:00:14,150 --> 00:00:16,520 to buy a nice shiny new Bitcoin mining rig. 6 00:00:18,760 --> 00:00:21,810 Now you expect that you will find a block 7 00:00:21,810 --> 00:00:24,020 in about 14 months with this fancy new rig. 8 00:00:25,040 --> 00:00:30,880 And remember that a block is worth about $15,0000 at today's prices. 9 00:00:32,260 --> 00:00:34,840 So if you amortize that you could say that the expected 10 00:00:34,840 --> 00:00:37,671 revenue of this box is about $1,000 per month. 11 00:00:39,500 --> 00:00:42,720 And now maybe once you factor in electricity and your other costs 12 00:00:42,720 --> 00:00:46,590 of operating it, if you actually got a check in the mail every month for $1,000. 13 00:00:46,590 --> 00:00:52,400 That would make it worth it for you to buy this $6,000 mining rig. 14 00:00:52,400 --> 00:00:55,010 But remember that mining is a random process. 15 00:00:55,010 --> 00:00:57,580 You don't know when you're going to find the next block. 16 00:00:57,580 --> 00:01:00,450 It's a completely random search. 17 00:01:00,450 --> 00:01:02,360 And you can find your next block at any time. 18 00:01:04,190 --> 00:01:09,020 So if we look at the distribution of how many blocks you're likely to 19 00:01:09,020 --> 00:01:13,660 find in the first year with that hypothetical one terahertz blocks, 20 00:01:14,940 --> 00:01:16,910 the variance is pretty high. 21 00:01:16,910 --> 00:01:19,740 Because the expected number of blocks you're going to find, is so low. 22 00:01:21,170 --> 00:01:26,190 So if we just look at the distribution, and this is a Poisson distribution, 23 00:01:26,190 --> 00:01:29,900 there's over a 40% chance that you won't find any blocks in the first year. 24 00:01:30,910 --> 00:01:34,850 In which case you might really be in trouble if you haven't earned any revenue 25 00:01:34,850 --> 00:01:38,720 in an entire year of running that $6,000 box that costs a lot of 26 00:01:38,720 --> 00:01:39,639 electricity to run. 27 00:01:41,950 --> 00:01:46,050 There's about a 36% chance that you'll find one block in the first year 28 00:01:46,050 --> 00:01:48,360 in which case maybe you'll barely scraping by. 29 00:01:48,360 --> 00:01:52,750 And then there's a slightly smaller chance that you'll find two or 30 00:01:52,750 --> 00:01:56,200 more blocks in which case you'll really be making a profit with this thing. 31 00:01:57,830 --> 00:02:01,400 So again, on expectation you might be 32 00:02:01,400 --> 00:02:05,280 just doing okay enough to make a return on your money. 33 00:02:05,280 --> 00:02:07,380 But there's a big chance that you'll make nothing at all. 34 00:02:08,720 --> 00:02:09,420 So that means that for 35 00:02:09,420 --> 00:02:13,640 a small miner, mining is essentially a big game of roulette. 36 00:02:17,010 --> 00:02:21,910 Now, historically when a small business people faced a lot of risk and they wanted 37 00:02:21,910 --> 00:02:26,680 to lower the risk, they got together and formed mutual insurance companies. 38 00:02:26,680 --> 00:02:29,280 So farmers would get together and 39 00:02:29,280 --> 00:02:33,030 agree that if any individual farmer's barn burned down, 40 00:02:33,030 --> 00:02:37,190 they would share profits with that farmer so that that farmer wouldn't go bankrupt. 41 00:02:39,360 --> 00:02:42,810 And the question is can we have a mutual insurance model that works for 42 00:02:42,810 --> 00:02:44,270 small Bitcoin miners? 43 00:02:46,430 --> 00:02:48,080 So that's what a Mining pool is. 44 00:02:49,200 --> 00:02:54,670 Now, the goal of a Mining pool is that a group of miners will get together, 45 00:02:54,670 --> 00:02:58,140 form a pool, and they'll all attempt to mine a block. 46 00:02:58,140 --> 00:03:03,470 Which pays the coin base, the newly minted coins to the same recipient. 47 00:03:03,470 --> 00:03:05,980 That recipient is going to be called the pool manager. 48 00:03:07,220 --> 00:03:11,440 So no matter who actually finds the block, the pool manager will control the rewards. 49 00:03:14,340 --> 00:03:16,980 And then the pool manager will take that revenue and 50 00:03:16,980 --> 00:03:19,990 distribute it to all of the participants in the pool 51 00:03:19,990 --> 00:03:22,960 based on how much work each participant actually output. 52 00:03:24,790 --> 00:03:27,820 Of course, the pool manager will also probably take some kind of cut for 53 00:03:27,820 --> 00:03:30,310 their service of managing the pool. 54 00:03:33,520 --> 00:03:34,830 So this sounds like a great idea. 55 00:03:34,830 --> 00:03:38,610 You can see why it would be attractive to BitCoin miners to join a pool to lower 56 00:03:38,610 --> 00:03:40,450 their variance. 57 00:03:40,450 --> 00:03:43,728 But how does the pool manager know how much work each of the members of the pool 58 00:03:43,728 --> 00:03:44,771 is actually performing? 59 00:03:45,820 --> 00:03:48,797 Obviously the pool manager doesn't want to just take everybody's word for 60 00:03:48,797 --> 00:03:51,587 it because everybody would say that they're performing more work than 61 00:03:51,587 --> 00:03:52,389 they actually did. 62 00:03:55,117 --> 00:03:57,978 So there's a really powerful idea which is that miners 63 00:03:57,978 --> 00:04:02,280 can prove probabilistically how much work they're doing by outputting shares. 64 00:04:02,280 --> 00:04:04,685 And shares are blocks that are almost valid. 65 00:04:06,104 --> 00:04:10,951 So whereas it's pretty rare to find an actually valid block that starts 66 00:04:10,951 --> 00:04:15,490 with all 66 required zeroes at the beginning of the output hash. 67 00:04:17,020 --> 00:04:21,453 There will be a lot more almost blocks that start with a lot of zeros but 68 00:04:21,453 --> 00:04:25,774 not quite the 66 necessary to make it a valid block. 69 00:04:25,774 --> 00:04:28,730 So a common choice might be say 40 bits or 70 00:04:28,730 --> 00:04:33,510 maybe 50 bits depending on what size the pool is designed for. 71 00:04:35,800 --> 00:04:40,420 Now if miners participating in the pool send in a bunch of these near valid 72 00:04:40,420 --> 00:04:43,980 blocks, the rate at which you're finding those near valid blocks should give 73 00:04:43,980 --> 00:04:49,990 the pool manager a very good statistical idea of who's doing how much work. 74 00:04:49,990 --> 00:04:53,620 And the other nice property of this is that there's no way to fake it. 75 00:04:53,620 --> 00:04:57,429 Because of the properties of the hash function, there's no way to find almost 76 00:04:57,429 --> 00:05:00,620 blocks without also finding actual blocks at the expected rate. 77 00:05:02,830 --> 00:05:04,230 So how does this whole setup look? 78 00:05:06,000 --> 00:05:11,290 So to start with, in each round after the previous block is discovered, 79 00:05:11,290 --> 00:05:17,275 the pool manager, which will be collecting transactions and assembling a block. 80 00:05:17,275 --> 00:05:19,450 Will tell all of the participants, 81 00:05:19,450 --> 00:05:21,470 here's the next block that we're all going to work on. 82 00:05:24,230 --> 00:05:27,140 So they'll assemble the block header. 83 00:05:27,140 --> 00:05:31,800 They'll assemble a Merkle tree of transactions to be included in that block. 84 00:05:31,800 --> 00:05:35,450 And in particular, they'll make sure to put into that mrkl_root of transactions, 85 00:05:35,450 --> 00:05:38,840 the coin based transaction which creates new coins and 86 00:05:38,840 --> 00:05:40,650 assigns ownership to the pool itself. 87 00:05:41,760 --> 00:05:46,240 And now this block header will be sent to all of the participants in the pool and 88 00:05:46,240 --> 00:05:47,640 this is what they're asked to work on. 89 00:05:47,640 --> 00:05:51,260 And they have to prove that they have been working on it by sending in shares, 90 00:05:51,260 --> 00:05:52,979 showing that they've been hashing this block. 91 00:05:55,560 --> 00:06:00,350 Now once that gets sent out, all of the miners go about and do their work. 92 00:06:00,350 --> 00:06:04,980 And they'll all slowly find almost valid blocks or shares until finally one of 93 00:06:04,980 --> 00:06:10,048 the miners, hopefully in the pool, finds a valid block that will then be published. 94 00:06:11,854 --> 00:06:15,592 Now after this happens, all of the participants will send in all of 95 00:06:15,592 --> 00:06:20,133 the shares that they found to the pool manager, which should be pretty efficient 96 00:06:20,133 --> 00:06:24,040 for the pool manager to verify are valid shares that they've found. 97 00:06:25,280 --> 00:06:28,410 And then the pool manager will take all of the revenue earned from that block that 98 00:06:28,410 --> 00:06:32,020 was published and distribute it back to the participants. 99 00:06:32,020 --> 00:06:33,780 Based on how much work they actually did. 100 00:06:35,170 --> 00:06:38,930 Now notice that this miner all the way on the right here who did happen to find 101 00:06:38,930 --> 00:06:43,500 the block, actually receives less revenue than that miner on the left there, 102 00:06:43,500 --> 00:06:47,360 who didn't find the ultimately valued block, but they did find more shares. 103 00:06:48,400 --> 00:06:49,550 So there's no bonus for 104 00:06:49,550 --> 00:06:52,370 this miner on the right here that they actually found a valid share. 105 00:06:53,590 --> 00:06:56,080 They would have been better off if they had just mined as an individual, 106 00:06:56,080 --> 00:07:00,820 because they would have gotten to keep all of the revenue from that block. 107 00:07:00,820 --> 00:07:03,450 But of course for the miner on the left they're sure 108 00:07:03,450 --> 00:07:07,210 going to be glad that they joined this pool, because they would've got nothing if 109 00:07:07,210 --> 00:07:10,330 they had just mined on their own, because they happened to not find a valid block. 110 00:07:12,790 --> 00:07:16,680 So there are a bunch of variations on this model of exactly how you determine based 111 00:07:16,680 --> 00:07:21,630 on shares submitted, how much money each miner should get in mining rewards. 112 00:07:21,630 --> 00:07:23,370 So a couple of common ones. 113 00:07:23,370 --> 00:07:27,790 There's the pay per share model, where the pool manager just announces, 114 00:07:27,790 --> 00:07:31,550 I'll pay a flat fee for every share above a certain difficulty that you're willing 115 00:07:31,550 --> 00:07:32,990 to send me based on this block. 116 00:07:34,530 --> 00:07:37,730 In some ways that's the best for miners, because that's a guarantee. 117 00:07:37,730 --> 00:07:40,770 Every time they find a share, they'll get a certain amount of money. 118 00:07:41,910 --> 00:07:45,550 The pool manager is essentially absorbing all of the risk in that scheme. 119 00:07:47,220 --> 00:07:49,850 Of course, as a result, usually with pay per share, 120 00:07:49,850 --> 00:07:52,665 you pay the highest transaction fee to the pool manager. 121 00:07:55,050 --> 00:07:57,490 Another way to do it is a proportional model, 122 00:07:57,490 --> 00:07:59,931 where instead of paying a flat fee per share, 123 00:07:59,931 --> 00:08:05,410 the amount per share depends on whether or not the pool actually found a valid block. 124 00:08:05,410 --> 00:08:10,620 So every time a valid block is found, the rewards from that block are distributed 125 00:08:10,620 --> 00:08:13,429 to the members proportional to how much work they actually did. 126 00:08:14,560 --> 00:08:17,800 So in this case, the miners still bear some risk, 127 00:08:17,800 --> 00:08:20,910 proportional to the risk of the pool in general. 128 00:08:23,010 --> 00:08:24,850 So if the pool is large enough, 129 00:08:24,850 --> 00:08:28,450 the variants of how often the pool finds blocks will be fairly low. 130 00:08:28,450 --> 00:08:31,430 Proportional can still be a good approach. 131 00:08:32,960 --> 00:08:35,790 And it will provide lower risk for the pool manager. 132 00:08:35,790 --> 00:08:39,138 And it gets around a problem that's inherent to pay per share, 133 00:08:39,138 --> 00:08:40,906 which is that with pay per share, 134 00:08:40,906 --> 00:08:45,370 the miners actually don't have any motivation to send in valid blocks. 135 00:08:45,370 --> 00:08:46,970 That don't have any motivation not to, 136 00:08:46,970 --> 00:08:52,200 but miners have the option when they find a valid block of just keeping it a secret. 137 00:08:52,200 --> 00:08:53,940 Now they're hurting the pool by doing that. 138 00:08:53,940 --> 00:08:56,120 So this is purely an act of vandalism. 139 00:08:57,120 --> 00:09:00,480 But they have no real motivation since they get a flat reward per share, 140 00:09:00,480 --> 00:09:02,780 to actually send in the valid blocks. 141 00:09:02,780 --> 00:09:05,950 Whereas with the proportional scheme, the miners certainly want to send in the valid 142 00:09:05,950 --> 00:09:08,910 blocks because that triggers revenue coming back to them. 143 00:09:11,050 --> 00:09:14,480 And a final variation that I think is interesting and was popular for a while. 144 00:09:15,890 --> 00:09:20,620 Is that the pool owner actually collects no fee but 145 00:09:20,620 --> 00:09:27,080 minors can't receive any revenue out until their balance is over one bitcoin. 146 00:09:27,080 --> 00:09:29,950 So basically that means the new entrants to the pool don't 147 00:09:29,950 --> 00:09:31,880 make any money for a while. 148 00:09:31,880 --> 00:09:35,210 And then they make money at an even rate without paying any 149 00:09:35,210 --> 00:09:36,860 fee to the mining pool manager. 150 00:09:39,700 --> 00:09:42,690 So essentially, once you're established in a pool with this approach, 151 00:09:42,690 --> 00:09:45,140 it's a great place to be, but they're harder to break into. 152 00:09:49,400 --> 00:09:51,070 So this has become pretty advanced now. 153 00:09:51,070 --> 00:09:54,630 There are quite a few protocols to run mining pools. 154 00:09:54,630 --> 00:09:57,350 And it's even been suggested that these mining pool protocols should be 155 00:09:57,350 --> 00:09:59,370 standardized as part of Bitcoin itself. 156 00:10:00,470 --> 00:10:04,160 So, just like there's a Bitcoin protocol for running the peer to peer network, 157 00:10:04,160 --> 00:10:08,600 these protocols are an API for communicating from the pool manager to all 158 00:10:08,600 --> 00:10:13,340 of the members, when to work on a new block, what that block should be. 159 00:10:13,340 --> 00:10:16,420 And for the miners to send back to the pool manager the shares that 160 00:10:16,420 --> 00:10:17,090 they're finding. 161 00:10:18,840 --> 00:10:21,720 Another reason why this is so important is that some mining 162 00:10:21,720 --> 00:10:24,700 hardware actually supports these protocols at the hardware level. 163 00:10:24,700 --> 00:10:29,410 Now this makes it very simple to buy a piece of mining hardware, 164 00:10:29,410 --> 00:10:34,140 plug in into the wall, both the electricity and your network connection. 165 00:10:34,140 --> 00:10:35,100 Choose a pool and 166 00:10:35,100 --> 00:10:38,120 then it will start immediately getting instructions from the pool. 167 00:10:38,120 --> 00:10:43,340 Mining and converting your electricity hopefully into money. 168 00:10:47,130 --> 00:10:51,250 So mining pools first started around 2010 actually way back 169 00:10:51,250 --> 00:10:55,330 in the graphics card era of Bitcoin mining, which is several generations ago. 170 00:10:55,330 --> 00:10:58,670 And they were instantly very popular, for obvious reasons, 171 00:10:58,670 --> 00:11:01,360 because they lowered the variance for the participating miners. 172 00:11:03,050 --> 00:11:07,940 So by 2014 the vast majority of all miners are mining through pools. 173 00:11:07,940 --> 00:11:10,010 Very few miners mine on their own anymore. 174 00:11:11,780 --> 00:11:15,830 And a very interesting thing happened in June of this year. 175 00:11:15,830 --> 00:11:19,850 Where the largest pool, called GHash.io, got so 176 00:11:19,850 --> 00:11:24,030 big it actually had over 50% of the entire capacity of the Bitcoin network. 177 00:11:24,030 --> 00:11:28,230 Now this is something that people had feared for a long time, but 178 00:11:28,230 --> 00:11:32,400 essentially GHash offered such a good deal to all the participating miners 179 00:11:32,400 --> 00:11:34,250 that everybody wanted to join. 180 00:11:36,180 --> 00:11:39,390 Now since then, GHash has gone down a little bit, partly be design. 181 00:11:39,390 --> 00:11:43,860 They've made their fees a little bit less attractive to try to get smaller. 182 00:11:43,860 --> 00:11:47,690 But you can see in this pie chart of the largest mining pools, 183 00:11:47,690 --> 00:11:49,330 they still have about a third of the power. 184 00:11:49,330 --> 00:11:53,200 And it's basically only two mining pools today that control about 185 00:11:53,200 --> 00:11:54,520 half of the power in the network. 186 00:11:54,520 --> 00:11:57,800 And then there's a nice smattering of other pools thrown in there. 187 00:12:00,980 --> 00:12:03,290 So are mining pools a good thing? 188 00:12:03,290 --> 00:12:08,560 Well the advantages of mining pools are that they make mining more predictable for 189 00:12:08,560 --> 00:12:11,180 the participants and they make it much easier for 190 00:12:11,180 --> 00:12:13,029 smaller miners to get involved in the game. 191 00:12:14,035 --> 00:12:18,330 Whereas the variance would simply make mining infeasible for economic reasons. 192 00:12:18,330 --> 00:12:20,390 If you didn't have mining pools. 193 00:12:20,390 --> 00:12:24,050 You can operate a much smaller mining rig profitably if you're part of a pool. 194 00:12:25,840 --> 00:12:30,290 Another advantage of mining pools is that since there's one central pool manager, 195 00:12:30,290 --> 00:12:32,790 whose sitting on the network and assembling blocks. 196 00:12:32,790 --> 00:12:35,630 It actually makes it easier to upgrade the network because by 197 00:12:35,630 --> 00:12:39,700 upgrading the software that the mining pool manager is running, 198 00:12:39,700 --> 00:12:43,550 that effectively updates the software that all of the pool members are running. 199 00:12:45,310 --> 00:12:47,630 The disadvantages of the mining pools of course, 200 00:12:47,630 --> 00:12:50,100 is that this leads to centralization. 201 00:12:50,100 --> 00:12:54,140 It's an open question of how much power the operators of a large mining pool 202 00:12:54,140 --> 00:12:54,720 actually have. 203 00:12:56,160 --> 00:13:00,380 Of course miners are fairly free in theory to move between pools as much as 204 00:13:00,380 --> 00:13:01,460 they like. 205 00:13:01,460 --> 00:13:05,350 If you're making profit on one pool, you'll likely make very similar profit in 206 00:13:05,350 --> 00:13:08,680 another pool because the fees that they pay out have converged. 207 00:13:08,680 --> 00:13:12,000 They all offer about the same deal today. 208 00:13:12,000 --> 00:13:15,970 But in practice miners don't switch very often, simply because they're lazy and 209 00:13:15,970 --> 00:13:18,730 it's easier to keep using the pool that they've already signed up for. 210 00:13:20,190 --> 00:13:24,240 And another disadvantage of mining pools is that it lowers the population 211 00:13:24,240 --> 00:13:28,610 of people actually running a fully validated bitcoin node. 212 00:13:28,610 --> 00:13:32,710 So previously small miners all had to run their own fully validating node, 213 00:13:32,710 --> 00:13:36,890 they all had to store the entire block chain, and validate every transaction. 214 00:13:36,890 --> 00:13:40,290 Now most miners offload that task to their pool manager. 215 00:13:41,320 --> 00:13:45,096 And this is one reason why a couple of lectures ago, I said the number of fully 216 00:13:45,096 --> 00:13:48,525 validating nodes may actually be going down in the Bitcoin network. 217 00:13:50,760 --> 00:13:53,697 So if you're concerned about this level of centralization, 218 00:13:53,697 --> 00:13:57,660 you might ask could we redesign the mining process so that we don't have any pools? 219 00:13:57,660 --> 00:13:59,890 So that everybody has to mine for themselves. 220 00:14:01,370 --> 00:14:05,248 And once again, that's going to be a very interesting topic that we'll talk about 221 00:14:05,248 --> 00:14:07,713 in our lecture on alternative mining approaches.