One of the best things about handicapping sports and especially college basketball, is that there is always more to learn and different ways to think about things. Unless you are getting every game right you can always do better – pick more winners and win more money. there are a lot of people out there looking at sports and statistics in a lot of interesting and very powerful ways. With a little creativity and determination you can use these statistics to give you a better understanding of different sports, and to help you find a bigger edge.
The group of statistics that has been particularly interesting to me recently is win distributions. These are done in different sports by a number of different sites. The specific details may differ from site to site, but the basics are the same. Using whatever statistical analysis methods are appropriate every game of the season is simulated a large number of times – like 10,000. With each simulation you will get a final record for every team in the league. Over the course of those simulations teams wills end up with some records a lot, some records rarely, and some records likely not at all. Those results will typically arrange themselves into some type of curve that shows you the most likely number of wins for a team along with the likely range of wins. As the season progresses the projections of the remaining games can be combined with the results of the games already played to create projections that should become more accurate because there are fewer unknowns.
These distributions are interesting just to look at, but they have some handicapping applications as well. Here are five ways that you could use win distributions as a handicapper:
Give you a quick sense at the start of the season of where teams are at – Unless you follow a league very closely it is very hard to have a good sense of every team in the league and what is expected of them during the season. That’s especially true if you are a fan of betting on college sports because of the number of teams. Win distributions can provide you a quick sense of how good NCAA basketball teams are expected to be. You can’t just rely on these projections blindly because they are only as good as the numbers used to develop them, but they will provide a better portrait of what is likely to happen then you might have without the projections.
Show you how teams are changing during the season – If you check in on the win projections every few games you can see how what has happened to date has impacted expectations. For example, if a college basketball team that wasn’t supposed to do much – maybe finish with 14-17 wins – gets off to a hot start – they start 10-3, for example – then the changes in the win distributions can be very useful to tell you what is going on. If the win distributions have improved significantly then you can tell that the start is better than expected at hat you are dealing with a team that may be somewhat better than it seemed they were. If the win distribution hasn’t really changed, though, then you can tell that the team just started the season with an easy schedule and they are likely to fall back to earth once things get tougher – like in conference play.
Give you a fast means of comparing teams – It can be hard to handicap a basketball game when you know one team very well, and don’t know anything about the other team. There are a few good ways to get a sense of how the teams compare, and win distributions are as good as any. By looking at how the distributions of the two teams compare you can see if one team is clearly better or worse than the others. For sports bettors, this is just a starting point, of course – a record doesn’t mean anything without some context – but this is a solid starting point.
Shows you how teams stack up in the conference – Teams in the same conference in college basketball play more than half of their schedules against common opponents. That means that the win distributions of the teams are going to be a useful way of comparing the teams in the conference and seeing how they stack up – which basketball teams are contenders and which are pretenders. This can be an especially useful comparison once non-conference play is mostly over and conference play is getting underway. By subtracting the number of wins a team already has from their win distribution you can get the expected conference wins, and that makes it very easy to get a sense of relative strength, which will help you in placing your basketball bets.
Comparing record to distribution can show teams that are deceptively good or bad – Win distributions can be a great way for college basketball handicappers to spot teams that aren’t nearly as good as they seem to be – something that can be useful heading into playoffs in any sport. A team that has a win total that falls on the far right side of their win distribution curve the they have overachieved compared to expectations, and they could be destined to fall to earth. That doesn’t mean you should automatically bet against them, but it does mean that you should look further to see if they really are overachieving and therefore vulnerable. The same would be true on the other side of the curve – a bsketball team over there has underachieved, and could be positioned to rise up and surprise in the postseason.