Posts
I then use this system to rank the players for all the major statistics on three different levels: their total stats for the season, their total stats when the game is tied or their team is trailing, and their total stats with runners in scoring position. This creates a total for each level, which I assign a different value to. I take the full point total for their total stats, I take 30% of their total for their tied/trailing stats, and I take 50% of their total for their stats with runners in scoring position. Then I add up all 3 totals and I come up with the table below, which shows their total point amount and their ranking in my MVP system:
This table is all the stats through August 5th and obviously this only includes hitters. Since hitters will get the majority of the MVP votes, I am excluding pitchers in this analysis. As you can see from the table, Josh Hamilton rates out as the current MVP of the AL. This is no surprise but you might be a little surprised to see A-Rod and Milton Bradley so low because they currently are the top two players in the AL in OPS. Bradley has 1.037 OPS and A-Rod's is .988.
Bradley ranks so low because has a low point total for his stats with runners in scoring position, since he has not had many at-bats in these situations. This is causing all of his totals(hits, home runs, RBI, etc) to be lower than the rest of the players and it is awarding him less points. He also has walked, been hit by pitch, or intentionally walked an incredible 49 times with RISP. He does get credit in my ranking system for walking more than the other players but it is not enough to make up for his low totals in other categories. For instance, he has only 3 HR and 30 RBI with RISP. He also is hitting only .270 with RISP, which is decent but the average of all the players is .297 so he is well below that and losing points there as well.
A-Rod ranks so low because he has been struggling with runners in scoring position and since he has a lower OPS than Bradley, it makes sense that he would be behind Bradley in my MVP Rankings as well. Here are A-Rod's stats with RISP through August, 5th:
.245 avg/4 HR/ 40 RBI/ .407 OBP/ .412 SLG/ .819 OPS/ 23 BB/ 32 K
Here are the average stats of all the players listed with RISP:
.297 avg/4.5 HR/ 45.7 RBI/.405 OBP/ .506 SLG/ .911 OPS/ 19.2 BB/ 20.6 K
As you can see A-Rod is below average in almost every category with RISP, except for walks and OBP. This is what is causing his point total to be so low and dropping him to 8th in my MVP rankings. Justin Morneau, Josh Hamilton, Ian Kinsler, and Carlos Quentin are 4 of the top 5 players with RISP in both leagues, so it is no surprise they rank so high in the MVP rankings.
Here are the total stats for the top 23 AL MVP candidates through August 5th:
Moving on to the National League, we see that Albert Pujols comes out as the MVP in my ranking system. He has the highest OPS in the NL at 1.078 and he has strong stats in all situations, so it is no surprise to me that he came out on top. It might be surprising to some though because Pujols is such a great hitter that we often take him for granted when it comes to MVP discussion. In my mind, if Pujols continues to hit this well the rest of the season and the Cardinals stay in contention for the playoffs, then Pujols should be the clear cut favorite for MVP.
There's a couple things that might jump out at you from this list because they definitely jumped out at me. Ryan Howard ranks pretty high for someone who is currently hitting .242 with 145 strike outs and Hanley Ramirez ranks extremely low but there is a good explanation for both of their rankings.
In Howard's case, he does have a high number of home runs and RBI, so he scores well in those areas but the reason he ranks as high as he does is because of his stats with RISP:
.326 avg/9 HR/ 66 RBI/ .442 OBP/ .591 SLG/ 1.033 OPS/ 29 BB
As you can see, Howard is hitting very well with RISP and coming through when the Phillies need him most. Howard ranks 4th overall with RISP and is 1st in the NL by a wide margin.
For Hanley Ramirez, the opposite is true because he has been awful this year with RISP:
.186 avg/5 HR/25 RBI/.333 OBP/ .395 SLG/ .728 OPS/19 BB
The fact that Ramirez has been so poor with RISP this year was the most shocking thing I learned from this whole analysis. He isn't even the worst hitter with RISP though. Jason Giambi is slightly worse because he is hitting only .206 with a .655 OPS.
Here are the total stats for the top 23 NL MVP candidates through August 5th:
Update: MLB MVP Race - Part 2 & MLB MVP Race - Albert Pujols & Kevin Youkilis
More: MLB
It's cool and interesting that you tried to create a new system for measuring value. However, I would suggest that if you take another stab at it in the future you should refrain from using things such as RISP.
ReplyDeleteUsing RISP, for one thing, is drawing from a small sample size. Not to mention the fact that Ramirez has had more than 30 fewer opportunites than the average player. Also, using this stat obviously doesn't account for all the production they bring when runners aren't in scoring position, which most of the time is just as important. I don't know how many extra points you award for being a shortstop, but it should be pretty significant given the importance of the position.
Any system that ranks Alfonso Soriano's season as being the 7th best in the league definitely has some major kinks in it. I recommend you model something a little closer to the VORP system on Baseball Prospectus. All the statistics they use really are very valuable and it currently ranks him at 39th in the National League--where he should be.
Winning percentage is another thing that should not be factored into the equation. Team success does not equal individual value. I'm not sure how large of role that played in Grady Sizemore being at sixth, but that's something that jumped out at me as a significant flaw. BP's VORP system has him first, which I agree with all things considered. A-Rod is right behind him in second.
I always enjoy reading someone's new idea or system when trying to evaluate a certain aspect of a player. I think you're on to something if you make a few specifications, because overall your top 23 in each league basically consists of players that should be there.
First off, thanks for the feedback.
ReplyDeleteI just wanted to clear up a few things as far as my system goes.
The system is my belief of what an MVP should be so it is a little different then just looking at a players value overall. I would agree that this model would not useful to predict overall player value.
However, it is supposed to predict how the MVP voting should go by not only considering overall stats by also stats in certain key situations as well as defense and winning percentage.
So in the case of Sizemore, yes he might be the most valuable in VORP, etc. but his team stinks and my model says that his stats are not that much better than the players ahead of him to be deserving of more MVP votes to make up for his team sucking. In my model, he actually loses 27points for being on a bad team. In comparison, Josh Hamilton gains 9 points for being on a decent team. I even gave Grady 5 extra points for defense.
Depending on the SS, it would change the amount of points awarded. I actually gave Hanley no extra points because he is horrible in the field. If someone played defense at SS like Ozzie Smith I would give them 5 points. Also, this number is totally based off my opinion, so it is somewhat flawed.
As far as the point break out goes, take for example Ryan Howard who has 92.07 points for his total stats, 79.05 points for his tied/trailing stats, and 144.66 points for his stats with RISP. To get his final total of 188.11, I took 100% of his points for total stats, 30% of his points for tied/trailing(23.71) and 50% of his points for RISP(72.33). So I am placing less value on the situational stats as compared to the total stats.
As far as the small sample size: Yes, it is small, which is why I didn't take the full amount, just a percentage of the total, because some guys just have less at bats than others but also for all these stats I have been using per at bat averages and assigning point totals to those too. So if someone is hitting 10 HR per at bat but only has 70 at bats then I give them credit for their per at bat averages, which I feel makes up for the low at bat total.
So to sum things up, this is an MVP ranking only, not total value per game, which is why I included some extras like RISP and Winning Percentage. In my mind, how you perform in key situations in a game is something that can make you the most valuable player. For instance, if you are hitting like an MVP then you should be driving runners home when they get into scoring position and making the other team pay. I feel I have to put some value on that if some players do it consistently better than others. Also, if you put up big numbers on a crappy team, then I don't think you should get as many MVP votes if other guys put up similar stats and did play for winning teams, which is why I factor in winning percentage. Otherwise, I would agree that RISP, Winning %, etc are not important to a players total value in general.
So if someone is hitting 10 HR per at bat but only has 70 at bats then I give them credit for their per at bat averages, which I feel makes up for the low at bat total.
ReplyDelete---------------------------------
This should say if someone has 1 HR every 10 at bats.
Also, in the case of Soriano his numbers overall rank him 13th in the NL but with RISP he is 2nd in the NL. In just 72 at bats he has 7 HR, 38 RBI and a 1.046 OPS. So he is hitting a HR every 10.2 at-bats and driving in a run every 1.9 at bats with RISP. The RBI/at bat is 8th best in both leagues. hamilton is first with an RBI every 1.6 at bats with RISP.
I would be interested in how well your system would play out past performances...like howard and pujols in 06?
ReplyDelete