FanNation - The Republic of Sport

For the last two months I have presented a number of rating systems, each of which values different aspects of a team???s performance: winning, scoring and offensive stats. The four rating systems are: W/L, which uses only wins and losses RS/RA, which uses runs scored and runs allowed Hybrid, which uses run differentials, except that each additional run counts half the previous one ??? it interpolates between W/L and RS/R Runs Created, which calculates an offensive and defensive rating using Bill James??? Runs Created formula (2005 version) Runs Created attempts to measure raw offensive and defensive strength, while the others are more results-based. Each system accounts for schedule strength and has a home field advantage factor. The RS/RA and Runs Created ratings also include a park factor. The three results-based rankings (W/L, Hybrid and RS/RA) at the end of the regular season, including the change from the previous rankings: Team W L W/L +/- Hybrid +/- RS/RA +/- RS RA Tam 97 65 1 +1 2 +1 3 774 671 Ana 100 62 2 -1 3 -1 7 -2 765 697 Bos 95 67 3 1 1 845 694 NYY 89 73 4 +1 5 +2 5 +3 789 727 ChW 89 74 5 -1 7 -2 4 811 729 Tor 86 76 6 4 2 714 610 ChC 97 64 7 6 6 +1 855 671 Min 88 75 8 8 8 -2 829 745 Phi 92 70 9 10 -1 10 799 680 Cle 81 81 10 9 +1 9 805 761 Mil 90 72 11 12 +1 13 +2 750 689 Tex 79 83 12 14 +3 17 901 967 Hou 86 75 13 +2 15 22 712 743 NYM 89 73 14 11 12 +1 799 715 StL 86 76 15 +2 13 +5 15 +1 779 725 Oak 75 86 16 -3 16 -4 14 -2 646 690 KCR 75 87 17 +3 18 +3 19 691 781 Det 74 88 18 -2 17 -3 11 821 857 Fla 84 77 19 20 20 770 767 Bal 68 93 20 -2 19 -3 16 -2 782 869 LAD 84 78 21 21 -2 18 700 648 Ari 82 80 22 22 21 720 706 Cin 74 88 23 24 25 704 800 Atl 72 90 24 +2 23 23 753 778 Sea 61 101 25 +2 25 24 671 811 Col 74 88 26 -2 26 26 747 822 SFG 72 90 27 -2 28 -1 29 -1 640 759 Pit 67 95 28 27 +1 27 735 884 SDP 63 99 29 29 28 +1 637 764 Was 59 102 30 30 30 641 825 Boston has lead in the RS/RA rankings since their introduction two months ago. The Red Sox have also lead in the Hybrid rankings for the last three weeks. Tampa Bay took over the top spot in the W/L rankings, passing Anaheim in the last week. The Cubs have always been the top NL team in all three ranking systems since they were introduced. In the AL, the W/L and Hybrid rankings agree with the actual playoff teams (although not necessarily on the AL East winner), while in the RS/RA rankings Toronto was favored over Tampa Bay for the wild card (both behind Boston). In the NL, all of the rankings agree with the actual division winners, but only the W/L ratings favored the Brewers over the Mets for the wild card. The Runs Created ratings at the end of the regular season: Rank +/- Team Rating W L RSPG RAPG ORC DRC 1 Bos .657 95 67 5.22 4.28 4.97 3.51 2 Tam .619 97 65 4.78 4.14 4.74 3.66 3 Tor .609 86 76 4.41 3.77 4.13 3.26 4 +1 NYY .586 89 73 4.87 4.49 4.70 3.91 5 -1 ChW .577 88 74 4.98 4.47 4.59 3.89 6 ChC .568 97 64 5.31 4.17 4.71 4.08 7 Ana .547 100 62 4.72 4.30 4.25 3.85 8 +1 Cle .540 81 81 4.97 4.70 4.43 4.07 9 +2 Tex .530 79 83 5.56 5.97 5.12 4.80 10 Min .527 88 74 5.09 4.57 4.75 4.48 11 -3 Det .526 74 88 5.07 5.29 4.68 4.42 12 Oak .508 75 86 4.01 4.29 3.83 3.76 13 Bal .506 68 93 4.86 5.40 4.56 4.50 14 +4 KCR .505 75 87 4.27 4.82 4.14 4.10 15 +2 LAD .501 84 78 4.32 4.00 4.32 4.32 16 -2 NYM .500 89 73 4.93 4.41 4.56 4.56 17 -2 Phi .500 92 70 4.93 4.20 4.37 4.37 18 -2 Mil .498 90 72 4.63 4.25 4.32 4.34 19 StL .496 86 76 4.81 4.48 4.75 4.79 20 Ari .485 82 80 4.44 4.36 3.80 3.92 21 +1 Fla .482 84 77 4.78 4.76 4.51 4.69 22 -1 Atl .479 72 90 4.65 4.80 4.09 4.28 23 Sea .449 61 101 4.14 5.01 4.11 4.58 24 Col .431 74 88 4.61 5.07 4.01 4.65 25 +1 Hou .428 86 75 4.42 4.61 3.79 4.42 26 -1 SDP .426 63 99 3.93 4.72 4.26 5.00 27 SFG .408 72 90 3.95 4.69 3.68 4.48 28 Cin .387 74 88 4.35 4.94 3.82 4.88 29 +1 Pit .376 67 95 4.54 5.46 4.09 5.35 30 -1 Was .366 59 102 3.98 5.12 3.61 4.83 Like the RS/RA ratings, Boston leads the RC ratings. Runs Created also agreed with the actual playoff teams, except for favoring the Mets over the Brewers (and the White Sox over the Twins). So what do all these ratings tell us? A list of observations verifying what we probably already knew: 1. Boston was the most powerful team in runs-scoring and raw stats, although it didn???t always translate into wins as it did for Tampa Bay and Anaheim. The Cubs were the class of the NL. 2. The AL East overall was by far the strongest division. The NL West overall was the weakest division. 3. The AL was much stronger than the NL, as the Cubs, the best team in the NL, ranked at best 6th or 7th overall. The AL was an overwhelming 149-103 in inter-league games, and all the ranking systems adjusted accordingly. 4. Comparing the W/L and RS/RA ratings, both Toronto and Detroit did not seem to convert their runs into wins, while Anaheim, Texas and Houston did so very efficiently. This is consistent with the difference between the actual wins and predicted wins using the Pythagorean formula that translates actual RS and RA into win percentage. Some not so obvious conclusions: 1. Toronto was rated surprisingly high, surpassing even the Rays in the RS/RA ratings. They may not be that far from seriously contending in the tough AL East. 2. Tampa Bay, thanks to its strength of schedule, was ranked higher than Anaheim even in the W/L ratings (Anaheim was rated as having the weakest schedule in the AL). 3. The Dodgers, thanks to a late season surge (thank you Manny!), now appear to be even with the Mets and Phillies for second-best in the NL as measured by their raw stats. Too bad the Brewers outdueled the Mets for the wild card ??? a Phillies-Dodgers division series might have been a much closer battle than indicated by their records. 4. According to the raw stats, the White Sox should have won the AL Central by a comfortable margin (by six games over Cleveland and eight over Minnesota), and Cleveland and Detroit should have given Minnesota a better fight. 5. Texas was not that far behind Anaheim in the strength of their raw stats, equivalent to three games over a full season instead of the actual 21 games back. Some teams rated much better in the RS/RA ratings than with RC, notably Minnesota, Philadelphia and Cincinnati, meaning they apparently did more (i.e., score runs) with less (fewer opportunities). Some of this difference could be due to random chance, but part of it might be due to the fact that RC does not take into account all aspects of the game, especially things such as base-running. The difference in expected wins per season was eight for Minnesota. Anaheim also is rated about eight wins better by RS/RA than by RC. Perhaps these teams did better at the ???intangibles??? not measured by the traditional stats. On the flip side, Texas and San Diego both were rated 6 expected wins worse by RS/RA than by RC. What can the rating systems tell us about the post-season? All but the Hybrid system calculate win percentages directly (in the W/L system) or indirectly via the Pythagenpat formula (RS/RA and RC). Therefore win probabilities for the playoffs can be calculated. These will be presented in a separate article.