[Chance Harper]

Angels and Cardinals: Second half fades?

With the first half of the 2008 MLB season coming to a close this weekend, it's time to take inventory of a few teams that just may be overachieving up to this point and could be in for disappointment ahead. While many point to the two surprises from the Sunshine State, the Rays and Marlins, the two teams using a lot of smoke and mirrors thus far are the Cardinals and Angels. Could one or both miss the playoffs?


Behold the evil Dr. Stat. Every Sunday, the Tampa Bay Rays show a superhero cartoon called 'Defenders of the Game' at Tropicana Field (you can also see the clips here on the official Rays website). This is Exhibit A in why handicappers can make so much money on the MLB betting odds. The league believes its fans need to be protected from...information.

Heaven forbid fans of the Rays should find out their team might not be as good as advertised. Tampa Bay is 55-35, but with a pythagorean record of 53-37. The Rays are 16-11 in one-run games, which makes the difference here. The 55-39 Boston Red Sox, meanwhile, are 13-16 in one-run games and have a pythagorean record of 57-37.

If you ignore the actual wins and losses, the Sox have outperformed the Rays and can be expected to catch up to Tampa Bay in the American League East. Ah, but that’s what those evil Red Sox would have you believe, with their stat-hugging front office and their two World Series in four years. Evil.

At least the Rays don’t have that much difference between their actual and pythagorean records. Fading them in the second half might not bring in a lot of cash for handicappers; the following two teams, on the other hand, are directly in the sights of my stat-fuelled satellite’s money ray. Cue the mad scientist music!

American League: Los Angeles Angels
The Angels have taken charge of the AL West at 54-37, 4˝ games ahead of an Oakland team that has just traded Rich Harden to the Cubs. But the A’s have a run differential of +62 compared to +23 for Los Angeles.

The Halos have a pythagorean record of 48-43, and they’ve already got nearly all their injured players back in the lineup – looks like Kelvim Escobar might be done for the year, so no help on the horizon there.

National League: St. Louis Cardinals
This was a tough call. The Florida Marlins have the biggest discrepancy in the NL between their actual record (47-44) and their pythagorean record (43-48). But they’ve already shaved off some of their betting value with 10 losses in their last 17 games, falling from first to third in the NL East.

They also have Anibal Sanchez (2.83 ERA, 5.20 xFIP last year) and Josh Johnson (3.10 ERA, 4.36 xFIP in 2006) due back in the second half. The 51-42 Cardinals (48-45 pythagorean) have further to fall, a tougher division to play in, and a shakier pitching situation.

Coming up this weekend...

Angels at Athletics
Game 1: Friday, 10:05 p.m.
LAA: Jon Garland
OAK: Sean Gallagher

Game 2: Saturday, 9:05 p.m.
LAA: Ervin Santana
OAK: Dana Eveland

Game 3: Sunday, 4:05 p.m.
LAA: Joe Saunders
OAK: Justin Duchscherer

The Harden trade left many baseball observers scratching their heads. Harden has a club option for the 2009 season, so he’s more than just a “rent-a-player” for the Cubs. And Oakland, however penny-pinching, is very much in the playoff hunt. But the team needs more hitting than pitching at this point. Oakland is 28th in team OPS at .696 and first in ERA at 3.40.

The problem for the A’s is that none of the three hitters involved in the trade are coming to Oakland right away – not even Matt Murton (career .810 OPS), who was sent to AAA-Sacramento. So the A’s will have to continue to get by on the strength of their pitching, led now by Dana Eveland (3.50 ERA, 4.67 xFIP) and Justin Duchscherer (1.78 ERA, 4.04 xFIP).

Gallagher isn’t chopped liver, though. He put up a 4.45 ERA and a 4.49 xFIP in 10 starts with the Cubs in his first full year in the bigs. None of the Angels have ever faced him at this level, so he’ll at least have the element of surprise on Friday. Jon Garland’s numbers aren’t Hardenesque, either: 3.76 ERA, 4.54 xFIP. Factor in the weak bats the Angels are swinging (.706 OPS, 25th overall), and Oakland will continue to hang tough in the AL West.

The Angels are -115 road faves at most books for Game 1 with a total of 8˝ runs.

[Capwizards]

I appreciate the write up but disagree.......Particularly on the Angels. Their pitching has been overacheiving to some extent, but they are still solid. Their team is very steady...they have an underacheiving lineup, good pitching, and a solid reliever. They also are a good road team which will help them down the stretch.....I would not fade them 2nd half. STL is a better option to keep an eye on. Aside from Loshe, their pitching has been inconsistant at best. Mulder isn't going to be the healthy pitcher they thought he would be.....They could be in trouble..

[EaglesPhan36]

Definitely see the Cards being so with the Cubs & Brewers acquisitions of the past week. Angels I think in a weak division are still the class. Wouldn't lay too much chalk on their lines - but I don't go over -140 or so anyhow.

[Willie Bee]

If you ignore the actual wins and losses...

I've gone around and around with Chance on the whole Pythagorean theory bit. Just think it's dangerous to put all your eggs in that basket when trying to analyze and predict future events based solely on that stat.

But I'm nothing short of befuddled as to how one can even think about 'ignoring actual wins and losses.' The Pythagorean theory simply spreads out runs scored and runs allowed real nice and even over x-number of games. Think about how often you hear some fan say something like, "Oh, if we only had a couple of last night's runs today." Too many other variables go into the equation, from how a manager might use his bullpen from one game to the next to giving some regular position player a night off.

The best recent example of why the Pythagorean doesn't work and shouldn't be the primary tool for analysis is the 2007 Seattle Mariners who won 88 games despite being outscored (by 19 runs) over the course of the season. They were able to do that because of a very solid bullpen, especially Putz who had a tremendous year closing out the close games for them. They won 27 one-run games and another 12 two-run games. So in 39 victories, they outscored their opponents by 51 runs. Seattle also lost 18 times by more than five runs, with 10 of those losses defeats of nine runs or more. In those 18 defeats, they were in the hole 146 runs. Just using those 57 games, they're down nearly 100 runs to Pythagorean and termed 'lucky' if you just go by the formula.

The Cardinals might be a good pick to fall back from this point forward. Their offense is in a bit of a funk right now, and I haven't heard anything recently about when or if Chris Carpenter might make it back.

The Angels have a good bullpen, certainly one that can keep them in the close games. Assuming they don't suffer any major injuries, they seem far less likely to me to drop off in the second half.

[MrX]

But I'm nothing short of befuddled as to how one can even think about 'ignoring actual wins and losses.'

There's no need to be befuddled, just look at the math. Midway through a season, Pythagorean is simply a better predictor of future winning % than the actual W/L record.

Is it possible to do even better than Pyth? Sure, but if you had to choose between Pyth and actual win/loss record to predict the rest of the season, the choice is clear.

Too many other variables go into the equation, from how a manager might use his bullpen from one game to the next to giving some regular position player a night off.

There are two things that cause a team's win% to differ from it's Pyth%. The first is luck. The second is a combination of factors such as those you mentioned including bullpen leverage and managerial decisions. Midway through a season, luck plays a greater role than the other factors. Toward the end of the season, things are a little more balanced and the actual win% becomes as good of a predictor as Pyth%.

The best recent example of why the Pythagorean doesn't work and shouldn't be the primary tool for analysis is the 2007 Seattle Mariners who won 88 games despite being outscored (by 19 runs) over the course of the season. They were able to do that because of a very solid bullpen, especially Putz who had a tremendous year closing out the close games for them. They won 27 one-run games and another 12 two-run games. So in 39 victories, they outscored their opponents by 51 runs. Seattle also lost 18 times by more than five runs, with 10 of those losses defeats of nine runs or more. In those 18 defeats, they were in the hole 146 runs. Just using those 57 games, they're down nearly 100 runs to Pythagorean and termed 'lucky' if you just go by the formula.

Now I'm the befuddled one. Why is this an example of Pyth not working and not an example of a team that got lucky? Especially when we look at the Mariners of 2008, isn't the evidence pointing to the latter?

[Willie Bee]

There's no need to be befuddled, just look at the math. Midway through a season, Pythagorean is simply a better predictor of future winning % than the actual W/L record. Is it possible to do even better than Pyth? Sure, but if you had to choose between Pyth and actual win/loss record to predict the rest of the season, the choice is clear.

I just think it's dangerous to rely on any single statistical column as your predictor. Is there value in Pythagorean? Sure. Is it all you need to look at? Certainly not.

Now I'm the befuddled one. Why is this an example of Pyth not working and not an example of a team that got lucky? Especially when we look at the Mariners of 2008, isn't the evidence pointing to the latter?

Where in Pythagorean did it tell me that Putz was going to spend the majority of the first half on the DL? Don't you think he could've made a difference in some of those early blown leads? Where in Pythagorean did it tell me Miguel Batista's ERA would be two runs higher than it was last season (which was right on par with his career ERA)?

The Mariners were coming into this season with two of the starting pitchers from 2007 -- Ramirez and Weaver -- gone from the rotation. Go back and check how many runs those two allowed last year to contribute to the total runs allowed by Seattle.

Now we also needed to subtract some of the offense that Jose Guillen provided as well. But we're talking about a team that was middle of the pack in runs scored last year, and is last in the AL up to this point this season. The bullpen has actually come around after a very shaky start, but with Bedard in and out of the rotation, and both Silva and Washburn sucking pond water, and with Hernandez also missing 3-4-5 starts, I fail to see where Pythagorean served as a good predictor for all that has gone wrong in Seattle.

[ryanXL977]

stl sucks and wont win 85

[MrX]

I just think it's dangerous to rely on any single statistical column as your predictor.

Absolutely agree. In fact, I don't use it at all. But, if all I had to choose between Win% and Pyth% to predict a teams future record, there would be no question that I'd choose Pyth%.

Where in Pythagorean did it tell me that Putz was going to spend the majority of the first half on the DL? Don't you think he could've made a difference in some of those early blown leads? Where in Pythagorean did it tell me Miguel Batista's ERA would be two runs higher than it was last season (which was right on par with his career ERA)?

The Mariners were coming into this season with two of the starting pitchers from 2007 -- Ramirez and Weaver -- gone from the rotation. Go back and check how many runs those two allowed last year to contribute to the total runs allowed by Seattle.

Now we also needed to subtract some of the offense that Jose Guillen provided as well. But we're talking about a team that was middle of the pack in runs scored last year, and is last in the AL up to this point this season. The bullpen has actually come around after a very shaky start, but with Bedard in and out of the rotation, and both Silva and Washburn sucking pond water, and with Hernandez also missing 3-4-5 starts, I fail to see where Pythagorean served as a good predictor for all that has gone wrong in Seattle.

You're right that my attempt to use this season as evidence was poor (especially since I already admitted that when looking at an entire season of data, Win% is almost as accurate as Pyth%).

Still, I don't believe that presenting last year's Mariners team is in any way discrediting Pythagorean. In order to do that you really need to show a correlation between differences in Pyth% and Win% and some of the factors you mentioned. It's not an easy task. It's a fact that some teams can be expected to outperform Pythagorean, but only by a small margin. When you see a team that greatly outperforms their Pyth%, the major factor is luck.

[Willie Bee]

You're right, X, about the Mariners not being a great example why Pythagorean doesn't work; in fact, there were many, Chance included, that predicted a downturn for Seattle this year because of their 2007 PT stats. I used the wrong words no doubt when I said they were a great example. What I should've said was they were a great example of why you can't depend on it within a given season, with 2007 being the season in question based on the fact they were in the red over the 57 games I mentioned (39 one- and two-run wins; 18 defeats of five runs or more).

In fact, just to show how wrong I was about them, I'm currently holding a futures ticket on them for 2008 My thinking was their bullpen would be about as good, their rotation would be improved with the subtraction of Ramirez and Weaver and the addition of Bedard, and the offense being about the same.

[seaborneq]

I don't care which of the theories work, as long as the Angels fade their way to "only" 91 wins.

[ryanXL977]

angels cannot hit
they will win betwene 90-94 though

[chipski]

no fade on angels or cards ..

[seaborneq]

Don't fade the Angels, too good

[jjgold]

Both Cards and Angels will be fine in the second half, they are solid teams. I think Tampa has weakened somewhat.

Yanks are now the team to beat though

[ICE-BLOOD]

angels will clinch division early and then play .500 ball

[seaborneq]

angels will clinch division early and then play .500 ball

I sure hope so.