The Other Worlds Shrine

Your place for discussion about RPGs, gaming, music, movies, anime, computers, sports, and any other stuff we care to talk about... 

  • sports predictor software

  • Somehow, we still tolerate each other. Eventually this will be the only forum left.
Somehow, we still tolerate each other. Eventually this will be the only forum left.
 #146975  by Don
 Mon Jun 07, 2010 7:15 pm
Today I checked out accuscore.com which is apparently the engine ESPN uses in their playoff simulator. So apparently you could get for $500 this software that's supposed make you more money when you go gamble or whatever (what else would you do with it?) So I looked at its prediction and it says it thinks Celtics are supposed to win 51% of the time. I think you get 1 to 3 payout for betting Celtics, which means if you're really sure this software is correct you can bet in a way that will guaranteed money (something like say $2 on celtics, $2 on Lakers would get more than $4 expected value just on the wins from Celtics). Yes of course nothing is perfect but if your odds predict 50/50 versus a Vegas line that's essentially 25/75, that's a guaranteed money making opportunity if you simply bet on both sides roughly equally. So why would you even need to try to sell this software? Gambling is not illegal, after all.

On the same site it also lists their track record http://accuscore.com/expert-trend-analysis. I'm not too familiar on all the ways you can bet stuff but it seems like if this software is so great why don't they just bet with their own money. Casinos runs with much smaller margins compared to what their claimed rates (like 1-2% edge on Blackjack) so if they got a 10%+ edge over casino, why not do it? And this is probably even more reliable than say asking an expert on stock market to actually buy/sell those stuff because it's hard to forsee if some country will collapse tomorrow and mess it up. On the other hand on sports you can be reasonably sure that the Flyers and the Blackhawks will probably play 7 games with mostly the same guys they started with. Yes that's not always true but certainly it's not nearly as volatile as the finance system.

Sometimes I wonder if a guy just runs NBA Live 2010 and give it a fancy name, would that really be distinguishable from all the other systems people use to predict these things? I'm not saying they're all useless, but predicting Lakers to get to Finals isn't exactly rocket science.
 #146976  by Don
 Mon Jun 07, 2010 7:27 pm
As a sidenote, I'm not assuming the casino is infallible and that their odds are the best, but clearly if they're usually wrong then they would no longer be in the business. Further, if someone is consistently better than the casino, then eventually that guy will have more money than the casino. It'd basically be like Back to the Future where Biff ends up owning the world or whatever since he's got the sports almanac. If you make 5% money per bet and win 100 times, that's 131 times your money. If you started with a $1000 investment, that'd be $131K, and 20 million after another 100 wins, and at that point you got enough money to win the casino if you keep up your winning ratio. Yes of course you're not going to get 100 bets right in a row but 5% on average seems pretty reasonable if you can truly beat the odds.
 #146979  by SineSwiper
 Mon Jun 07, 2010 10:57 pm
Don wrote:but clearly if they're usually wrong then they would no longer be in the business.
That's a false assumption. There are plenty of companies making millions of dollars on vaporware.

You could also have a product that is mostly correct, but the margin of error would even things out. In other words, if you used this thing every time, you would get a net loss. Sure, you would get a lot of money for some of the bets, but if you added everything up, it would be a net loss. People who buy the thing will think it's awesome because of the money they got, not the ones they lost. Most gamblers aren't math majors, else they wouldn't need something like this.

There is the chance that this software is indeed beating the odds because the majority of people are not using the software (or something like it). For example, I always get a net gain when I play poker at the casino. Why? Because a majority of the people do not know how to play properly, and you are stealing money from those people. Even some of the good players end up playing aggressively, which you can counter easily. There are some good players out there (mostly retired folk who have been playing it 20-40 hours a week; shit, I would if I was retired), but if you play your cards right (literally), you will come out ahead.
 #146981  by Don
 Mon Jun 07, 2010 11:34 pm
Playing against other players is a zero sum game so obviously someone have to win (minus whatever the dealer rakes). On the other hand in a sports bar the house doesn't have to give equal odds. For example if two teams are really 50/50 they're not obliged to give you even money on either team. I think in that case it'd be $0.95 back on $1 wagered for either team, which is probably comparable to a typical casino game (5%).

But, these guys are claiming they have considerably better than 5% better than average (if you just guessed randomly you should come close to 50% winning rate when betting on two teams) which means it ought to be enough to overpower the house's advantage. I know they set limits and stuff but again, if you're sure you've a consistent edge over the house you'll eventually be able to beat them ala Back to the Future. Sure it'd take longer when you're not 100% but eventually you'll beat them. Yes the house can break even or make money for a while by making money from the guys who aren't smart betters, but eventually the guy who is winning consistently would have so much money that his bets become a significant portion of the total bet. Also, these software doesn't require any training or anything. You obviously just run it and it says what it thinks about who will win. So like even if they try to cap your winning you can certainly call up a friend and say 'yeah I got this software that's always guaranteed to make money, just bet this and that!' Therefore if you're certain your software is correct, you can easily put casinos out of business (and make money in the process), and there's no way they can make that up given the cost of passing that knowledge on is nothing.

So I see how no way these software is supposed to actually be competitive in a betting environment. Maybe this stuff is very useful for Fantasy Sports, but I mean if you can pay $500 to buy some software that's guaranteed to win, what's stopping the casino from doing hte same? And if they know the odds your software gives (which they can, since they can buy the same thing) they can set the odds in a way that it is impossible for you to win anyway.

BTW when I said 'if they're wrong they'll be out of business', I mean the sports bar itself, because obviously if I'm running around giving people 1 to 10 odds for Celtics to win, I'd end up losing most of the time. Yes I know these guys can lose money too, but if they're regularly losing money they won't be able to run their operation anymore!