Sunday, May 11, 2008

The Connections Between Stats and what they Tell Us

There are some some stats that are linked to other stats and therefore telling us the true outcome of that stat. I think it would be interesting if I go through most of them. The first being the connection between strikeouts and batting average. Although batting average has a 43% chance of changing from year to years it's been proven that people with a lot of strikeouts tend to have lower averages annually while players that don't strike out as much tend to be less predictable with some years having a high average while other years a low one. What this means is that a guy like Carlos Beltran can have his average sore to the .300's one year(2003) but have it drop dramatically the next year to the .260's(2004). A guy like Adam Dunn on the other hand, has a career batting average of .247 and has never hit higher than .266 but if you look at his average year by year they're pretty consistent because of the fact that he strikes out once every 3.5 ab's. As you see the lower amount of strikeouts you have the better chance your average could go up and down while if you strike out a lot you'll consistently have a low batting average. The reason for this is that the less you strike out the more chances you have to get a hit because the best thing you can do is to put the ball in play.

A hitter that hits a lot of doubles but not a lot of homers one season can expect an increase of home runs and a decrease in doubles the next. The reason for this is that both doubles and homers are symbols of power and therefore if you just miss them one year the next year you have a much better chance of getting that extra 5-10 feet and therefore more homers. This also works with walk rate and ISO(SLG.-AVG.=Isolated Slugging). If you have a high walk rate that means you take pitchers deep into counts so there is a bigger chance that the pitcher will then throw you a good pitch after you've fouled back 4 straight and have tired him out, thus increasing ISO which rates your pure slugging ability. A player that has a high walk rate and a middling ISO can expect an increase in ISO the following year becasue of the amount of pitches he sees and vice versa, if a player has a high ISO, meaning he hits many eXtra Base Hits, his walk rate will go up the following year in fear of getting pounded, a la Barry Bonds after his 73 homer season.

Now to go to newer stats. If hitters have a low BABIP(batting average on balls in play) you can expect their batting average to go up because your BABIP should always be at the same level as your BA. If his BA is low and you know they aren't career .240 hitters then they've just been having bad luck hitting balls in play and therefore you know it'll change soon. The same thing goes for LD%(line drive rate) and hits. If a player has a high LD% but isn't getting as many hits(AVG) as you would expect him to then you know his BA and hit rate will go up. Now, with a stat called HR/FB you can see if a power hitter is finished or is just having bad luck. If a guy that usually hits 30-40 homers a season isn't hitting them out, you could look if he's hitting it to the outfield and it just isn't getting that last push(Carlos Beltran) or if he's just stop hitting balls to the outfield(Frank Thomas).

Pitchers have peripheral stats like K/BB, K/9, BB/9, HR/9 and BABIP(used in the opposite way) to help us determine if a player is having good luck(Cliff Lee), bad luck(Jonny Cueto) or is pitching at his best. If a pitcher has a low ERA and bad peripherals or a great BABIP you know he won't keep up his performance, but if a guy has a 3.50 ERA and great peripherals you know he'll keep it up. How does all this help us predict? Because you can analyze a player's performance and then use all these methods to see if he'll continue to have success or if he's just a fluke. You can also use them to see if he'll have another good year or drop off or have a breakout year. I thought this would be an extremely cool post idea, so here it is. Tell me if you enjoyed and learned anything from this post.