Monday, April 27, 2009

After Twenty Games

My Statistical Analysis After 20 Games

Defense - Overall

The Jays have arguably the best defense in the AL, having the highest Defensive Efficiency at 0.719, ahead of Oakland at 0.717, kc at 0.710 and TB at 0.709. (Defensive Efficiency is the percentage of balls in play turned into outs.) The league average is 0.694. This means that if there were no strikeouts and only balls in play, the league average requires 39 outs to get through a game compared to the Jays 37.6 outs

Defense - Fielding

The Jays are at or near the top in almost every fielding category. They lead the league in fielding percentage at 0.990 and with the most chances at 778 and are second to Oakland in chances per game. They have also turned the second most DPs behind Cleveland

Defense - Pitching

Only KC (3.56) and Seattle (3.95) have allowed fewer runs per game than the Jays (4.25), well below the league average (5.10) which is up nearly one-half a run per game from last year's 4.68. The team ERAs are KC: 3.47, SEA: 3.50 Jays: 3.90, well below the league average of 4.83. The Jays are second in WHIP at 1.262 behind, KC at 1.209.

In the first 20 games, the Jays have 10 quality starts, although 12 times the starting pitcher has ended the 6th inning having given up three or fewer runs. The bullpen has been effective 13 of 20 games, with effective meaning they've only allowed half an earned run per inning pitched.

I have a metric that I prefer to measure the effectiveness of a pitching performance, batters faced per out (BF/O), which correlates well with WHIP but factors in aspects of defense (errors and double plays) as well as pitching style, as ground ball pitchers are often able to erase runners with double plays. A WHIP of 1.0 would correspond to a BF/O of 1.333 (=4/3) while the league average WHIP of 1.4 corresponds to a BF/O of 4.4/3 = 1.467. I use a BF/O of 1.4 as the standard of a pitcher being effective, which corresponds to a WHIP of 1.2. BF/O gives me an appreciation of how much trouble a pitcher is facing with runners on base. (BJ Ryan had BF/O if 1.4 or more 38.3% of the times he pitched last season compared to Rivera 15.6%, Nathan 22.1%, Soria 23.8% and Papelbon 31.3%. FRod was somewhat shaky too at 35.5% of the time having a BF/O >1.4.)

On this measure, Roy Halladay and Ricky Romero have been 100% effective, accounting for 8 of the 11 effective starts the Jays have had, with Purcey, Richmond and Tallet having one each. Out of the bullpen, Jason Frasor has been effective 100% of the time and Scott Downs 90%. In the next tier Murphy, Carlson, Camp, League, Tallet and Ryan have been effective between 50% and 67% of the time. Only Bullington, in his two appearances, has yet to be effective.


My preferred metric for offensive effectiveness is to convert runs created (RC) into RC/pa. A league average WHIP of 1.4 implies about 40 at bats per game and with the AL averaging 4.68 runs/game in 2008, it means that the team must average nearly 0.12 runs per at bat to achieve the league average. This metric correlates extremely well with OPS, but has the advantage of measuring offensive production in runs.

On this metric, the Jays have been doing very well. The team leader is Aaron Hill at 0.225, followed by Lind and Millar (0.191), Snider (0.182), Bautista (0.177), Overbay (0.166), Scutaro (0.157), Rolen (0.154), Wells (0.146) and Barajas (0.140). The only regular who is under-performing is Alex Rios, at 0.101, but his numbers have been improving lately. The first base platoon of Overbay and Millar have an RC/pa of 0.175 and the LF platoon of Snider and Bautista have a RC/pa of 0.180, although this is based on all Bautista’s plate appearances, not just when he has played LF. (Note that I have not made any adjustment for position played. The conventional wisdom is that more run production is expected from corner infielders and outfielders whereas most teams will concede some runs production up the middle for stronger defense.)


After a 14-6 start, the Jays have the best record in the AL. They have been hobbled at pitching, and the next 20-30 games could be quite a test. Tallet will need to continue to fill a hole in the rotation effectively and Richmond must continue to battle to succeed despite allowing too many runners. Purcey has great stuff but needs to last longer and allow fewer base runners to be effective in the rotation. One start is not enough to gauge Burres. By the end of May we may have seen some new and returning arms in the rotation. The bullpen is solid, and is improved with Ryan on the DL. The offence should continue to produce a lot of runs, but don’t expect them to continue to continue to average over 6 runs per game, although I think this squad will average over 5.5. Over the next 21 games they have 9 road games including a much-improved KC and an always-tough 5-day west coast swing through LA and Oakland. At home, they host Baltimore (3), Cleveland (2), the Yankees (3) and the White Sox (4). I realistically see them going 13-8 or 12-9 in this stretch.

Season Projection:

I’m a firm believer in “A win in April is as important as a win in September”, although that’s not to minimize Cito’s philosophy of sacrificing a win now for 3 or 4 down the road. It’s great to have 14 wins in the bank after 20 games. I doubt many expect the Jays to maintain a 0.700 pace for the rest of the season. Under Cito in 2008, the Jays were a 0.580 team. The offence is improved with the platoon of Snider and Bautista in left and the return of Hill at 2B. While the pitching has taken some hits from injuries, they’ve managed to get the job done so far with several pitchers expected to return in the next month. I went into the season projecting that the Jays would contend in September, playing at about the same 0.580 pace (94-68) they did last year under Cito. A 0.580 pace the rest of the way projects to a 96-66 record. Even if they slip back to 0.560 the rest of the way, they’ll finish with a 0.580 (94-68) overall.


  1. I am going to revamp my database over the next week or so, which will include some changes to the computation of RC. It seems that I have a flaw somewhere in a formula, as my values are slightly off compared to some other reliable sources.

  2. Some nice work here! There is another Jays fan working with numbers - take a gander over to

  3. I scanned through your blog and things do look interesting. I'm kind of with Christopher about the IBB, though. The treatment of the IBB is controversial. While it is putting a runner, and potential run on base, it is also taking the bat out of the hands of what is often a dangerous hitter.

    Anyway, I'm now following your blog and will comment more fully later.

  4. I realized after I posted my comment that you are not the Impact Stat guy.