How many runs does a player contribute to his team’s offense? One way to answer this question is with the statistic “runs created”. Runs created (RC) is not the most popular sabermetric measure today but it was one of the statistics which made Bill James famous. I discussed the calculation of the runs created formula in my recent article on offensive efficiency. The same version of the formula will be used for players except that it will be adjusted for ballpark.
Does RC work? As shown in the efficiency article, the team runs created comes close to the teams runs scored in most cases. Similarly, if you add the runs created for individual players on a team, it will also usually come fairly close to the team runs. This is an indication that runs created is doing a pretty good job of measuring what it is intended to measure: how much each player contributes to his team's runs scored total.
Another statistic is runs created per game or runs created per 27 outs (RC/G). Theoretically, this statistics tells you how many runs your team would score per game if you had the same player bat in all line-up positions. For example, Magglio Ordonez had an RC/G of 6.0 so you would expect a team of 9 Ordonez's to score 6 runs per game. That may not be a very practical or realistic use of the statistic. However, it’s a good statistic for comparing the relative offensive contribution of different players.
A player like Edgar Renteria who played a lot of games will have more runs created than a player like Matt Joyce who played only semi-regularly for part of the season. On the other hand,
As mentioned above, I adjusted RC and RC/G for the impact of the hitter’s home park (according to US Patriot). A hitter playing his home games in a hitter friendly park like Cellular Field in Chicago
You might be surprised that Comerica Park has a neutral park for run scoring over the past five years and thus no adjustments are made for the Tigers. Other American League park factors range from 108 (8% more runs than the average park) to Safeco (6% below average).
Table 1: Runs Created by Tigers in 2008
PA | RC | Lge Rank | |
Cabrera | 684 | 108 | 8 |
Granderson | 629 | 100 | 15 |
Ordonez | 623 | 94 | 22 |
Polanco | 629 | 85 | 34 |
Guillen | 489 | 69 | 53 |
Renteria | 547 | 58 | 77 |
Sheffield | 482 | 52 | 90 |
Thames | 342 | 44 | 107 |
Rodriguez | 328 | 42 | 116 |
Inge | 407 | 40 | 119 |
Joyce | 277 | 39 | *** |
Santiago | 156 | 26 | *** |
Raburn | 199 | 20 | *** |
Thomas | 133 | 18 | *** |
Larish | 111 | 11 | *** |
Ryan | 50 | 9 | *** |
Hessman | 31 | 8 | *** |
Jones | 90 | 4 | *** |
Hollimon | 25 | 4 | *** |
Clevlen | 28 | 2 | *** |
Sardinha | 49 | 2 | *** |
Table 2 ranks the Tigers among 136 American League players with 300 or more plate appearances in runs created per game. This table looks a little different because players like Mike Hessman, Dusty Ryan and Ramon Santiago produced more when they did play than some of the players with more plate appearances. Of those that played regularly, Granderson topped the list with 6.5 runs created followed by Cabrera (6.3), Ordonez (6.0) and Guillen (5.9).
Granderson, Cabrera and Ordonez did well and Guillen did OK when healthy but it was a far cry from 2007 when Ordonez created an amazing 9.3 RC/G and Polanco and Granderson produced over 7 per game. They were also not helped by the fact that three regulars or semi-regulars were at 4 RC/G or lower - Gary Sheffield, Edgar Renteria and Brandon Inge. They were helped by some nice performances off the bench by Matt Joyce, Clete Thomas, Ramon Santiago, etc.
For next year, we can hope for small improvements from Granderson, Cabrera, Inge and Renteria (if he is still around). Sheffield could also give them a boost if healthy but I wouldn't bank on it. There is also potential for declines with Polanco, Guillen and Ordonez getting older. They will probably need contributions from youth (Joyce, Ryan, etc) and new players (such as whomever comes to play catcher and shortstop) if they are going maintain or improve offensively next year.
Player | PA | RC/G | Lge Rank |
Hessman | 31 | 11.6 | *** |
Ryan | 50 | 7.5 | *** |
Santiago | 156 | 7.3 | *** |
Granderson | 629 | 6.5 | 17 |
Cabrera | 684 | 6.3 | 21 |
Ordonez | 623 | 6.0 | 32 |
Guillen | 489 | 5.9 | 37 |
Thomas | 133 | 5.7 | *** |
Joyce | 277 | 5.6 | *** |
Hollimon | 25 | 5.5 | *** |
Polanco | 629 | 5.4 | 46 |
Rodriguez | 328 | 4.9 | 63 |
Thames | 342 | 4.7 | 76 |
Sheffield | 482 | 4.0 | 109 |
Renteria | 547 | 4.0 | 114 |
Raburn | 199 | 3.7 | *** |
Inge | 407 | 3.7 | 121 |
Larish | 111 | 3.7 | *** |
Clevlen | 28 | 2.5 | *** |
Jones | 90 | 1.5 | *** |
Sardinha | 49 | 1.1 | *** |
I was surprised to see Hessman, although in just a handful of PA, at the top of the list.
ReplyDeleteTaking into account his minor league success why is the talk soley focussed upon Inge moving back to third instead of giving Hessman a legitimate shot at the starting role?
That's a good question because he can field too. I think the answer is they have money committed to Inge. If he hits and fields like he did between 2004-2006, then I want him at 3B. If he bats .200 in April though, I hope they explore other avenues and Hessman might be a decent option. They are also experimenting with Larish at 3B.
ReplyDeleteLee
Lee,
ReplyDeleteNice post. I really enjoy your number cruching. Runs produced per game really gives a nice quanitative approach to a player's offensive value or lack there of. Really shows the lack of production out of Sheffield, as he plays a purely offensive position (DH). Bet his number is lower than any other DH in league.
Arlie
Sheffield is not at the bottom but is definitely below average. The positional analysis is coming soon.
ReplyDeleteI had wondered earlier in the season, with the problems that Guillen was having, physically, and with adjusting to 3B intially, why they didn't try Hessman then.
ReplyDeleteHere's a question for you. Was the league OPS down this year from the previous year? If so, will it likely rebound next year, and in so doing would Guillen's (as well as everyone else you mentioned) numbers actually not decline next year, but he may decline in total rank. I hope you follow.
Have you done an analysis as to RC's over the course of the season, vs. Runs prevented, for the proposition of Inge vs. Guillen and 3rd for next year.
First, I'll address the runs saved/lost question. I did a quick and dirty analyses at motownsports of a switch of Inge to third and the addition of a defensive SS:
ReplyDeletehttp://www.motownsports.com/forums/
showpost.php?p=1653055&postcount=179
Based on a lot of assumptions, it came out to 12 runs better overall with Inge 3B/Jack Wilson SS/Guillen LF than with Guillen 3B/Renteria SS/joyce - Thames LF
As for the other question about the drop in offense, it was way down in the AL early in the season but it caught up later. It was still down at the end but not by a huge amount- between 0.1 and 0.2 runs per game.
ReplyDeleteI think it was down because the league is getting older. In the Tigers case, they also have a few players who are getting up there. I do think there is a better chance that Guillen/Polanco/Ordonez decline rather than get better. And it would be their raw numbers as well as their ranks.
Lee
Thanks. I'll look at he quick and dirty at motown.
ReplyDelete