Good sports handicappers know that getting an insight into how a game is likely to turn out requires a careful analysis of statistics that matter. Figuring out which of the statistics actually matter is a big issue, but one that we won’t deal with here. What I do want to look at today is situations in which bettors are unintentionally paying too much attention to a particular type of stat. Without knowing they are doing it handicappers can easily count statistics twice, and that can skew their view of games and lead them to make frustratingly inaccurate decisions. The easiest way for sports bettors to understand what I am talking about is with a couple of examples:
When a smart sports handicapper is trying to get a sense of a football game they are obviously going to want to have a sense of the comparative strengths of the quarterbacks. One of the more effective ways to get a quick sense of the pivots is by looking at the powerful Yards per Attempt stat. If you wanted more of a sense beyond that then some people may want to look at the QB rating. On the surface this could make sense – both are ways to measure how effective quarterbacks have been in a simple, easy to compare number. The problem, though, is that the quarterback rating is determined using a fairly complex formula, and part of that formula includes taking the total number of passing yards over a given period and dividing it by the number of passing attempts – the exact formula for finding the YPA. If you were using both of these stats, then, you would be counting their YPA twice. To many football handicappers, a player that had a particularly good YPA would look better than they should, and one who struggles with YPA would look weaker than they should and you may overlook other types of strengths as a result.
This seems to me like a blatantly obvious example, but I include it because I just saw someone make this exact mistake when talking about the World Series. When making the case that Texas had much better hitting than San Francisco they listed off a number of stats to make their case – they had a higher team batting average, a higher on-base percentage, a higher slugging percentage and higher OPS, and so on during the regular season. The problem is that OPS is short for on-base plus slugging, and all you do to calculate it is add together the on-base percentage and the slugging percentage. In other words, if Texas had a higher slugging percentage and on-base percentage than San Francisco then they automatically had a higher OPS. In the mind of the sports handicapper, Texas had an edge because of a long list of stats, but some of those stats were duplicates. There were fewer arguments for the Rangers than the writer actually gave them credit for. What he was saying was that he liked the Rangers better because they get on base more often than the Giants, they gain more bases per at-bat on average, and the get on base more often and gain more bases per at bat. That’s the definition of redundant.
Both of these situations can be problems by themselves. Where they can really create issues, though, is when you incorporate issues like this into some type of formula – like when you are calculating your own power rating. If your baseball power rating formula included on-base percentage, slugging percentage and OPS then you would be counting two stats twice, and you would be giving teams that can hit well disproportionately strong ratings. That might result in sports bettors losing more games than they should where an offense doesn’t meet expectations.
So, how do we avoid situations like this when we are double counting some stats? Here are three tips to help you out:
When possible use base stats – The easiest way for bettors to avoid double counting is by using the basic stats whenever possible. Instead of using OPS in a formula, for example, use slugging percentage and on-base-percentage. Instead of using ERA you could use earned runs allowed and number of innings pitched. When you reduce everything down to the base you can avoid unintentional errors, and you can be sure you are actually gathering information that is relevant to you. It also makes it easier for sports handicappers to make small tweaks to their formulas.
Understand the stats you are using – People talk about a lot of stats that they don’t actually understand. QB rating is a good example of that. People will happily tell you that t is a flawed statistic, but very few of those people actually know how it is calculated, so they don’t really know why it is flawed or how it could be improved. Taking the time to understand the stats you are using can help you avoid errors, and can make you more effective as well because you actually understand what it is you are using. This is ever more important now as math geeks are increasing ever more complex and valuable statistics that can be very effective if used properly, but can create nightmares for bettors who don’t really understand them.
Get a picture of what you want first, then choose your stats – Sports handicappers who don’t understand stats very well will just grab the stats that they like and then try to figure out what to do with them. It’s much more effective to first think about what you want to find out, and then to use that to figure out the best stats to use. For example, let’s say you want to know how good a defensive line is because you value the importance of good defensive line play. Once you know that you can decide whether you want to use sacks as a measure, of if you want to combine sacks and negative plays, or if you want to give the line credit for interceptions as well because pressure usually causes those, or so on. Stats are just tools to let sports handicappers paint the picture you want to paint, and then to compare those pictures of different teams in meaningful ways.