Seattle Mariners - an exception to Sabermetrics?

[TuesdaysWithRillo]

http://msn.foxsports.com/mlb/story/7110738

 

I found this article very interesting. Could "clutch hitting" actually have some meaning? It sure seems to for this team. The Mariners aren't getting on base, they aren't hitting for power, and their starting pitching is horrible, but their bullpen shuts the door, and they get clutch hits. What do you stat gurus think of this?

[homer]

First off, there are exceptions to rules all the time. But, the Mariners do pose an interesting case study if the trend holds.

Their Pythagorean record is 59-56 so they are about 6 wins ahead of that pace which isn't huge by any means but it's probably significant. Basically means they get lucky more than they are unlucky. That trend may hold but if I were a betting man, I wouldn't cash in the farm on it continuing.

[TuesdaysWithRillo]

I thought this part of the article was amusing:

 

Sabermetricians hate the word "clutch," and many dismiss it altogether, claiming it simply doesn't exist. Well, it does in Seattle.

[endaround]

Short answer not really. Long answer, there are teams that vary from their talent by a great degree each year. Seattle is one of those teams this year. They are as far away from their Pythogrean record as the White Sox but few talk about how lucky the Sox are. The Cards are in the same boat. So are the Nationals. Seattle just has a higher base talent level. They are a .500 team getting lucky rather than a .400 team getting lucky.

[crewcrazy]

The Arizona Diamondbacks are probably another example of a statistical anomaly this year (pyth record of 57-61, yet they're leading the West). It'll be interesting to see if either team can keep it up. Both teams are examples of how much having a great bullpen can help you. I've been waiting for Putz and Valverde to suffer their own respective slumps, but they haven't happened yet...both guys have been nearly automatic this year. It wouldn't surprise me if Putz got a couple Cy Young votes if the M's find their way into the playoffs.

[Brian the Automator]

I thought this part of the article was amusing:

 

Sabermetricians hate the word "clutch," and many dismiss it altogether, claiming it simply doesn't exist. Well, it does in Seattle.

I don't hate the word "clutch". Although, I do think it's inappropriately applied to small sets of data which doesn't prove anyway. Surely, as in any line of work, there are people who excel in the face of adversity, and those who fall apart. Sabermetrics simply believe that it's very difficult to predict which person is going to be which person.

 

[brettac1]

This should be about the time they fall apart, then. The Nats had a bunch of these articles written about them a couple years ago, and promptly regressed to the mean.

 

There's all kinds of things that can skew a pythagorean W/L, too. Namely getting blown out a disproportionate number of times (as will surely happen with the likes of Horacio Ramirez in your rotation), or vice versa. Great bullpens make a difference when you have a weak offense and weak rotation.

 

Even so, they success they have had doesn't really do anything to "disprove" sabermetrics. It's not like they said "we are going to acquire all the guys that we consider to be clutch". It's practically the same lineup they trotted out there last year that hit .265 with RISP and is now at .287.

 

In short, it's a fluke.

[mothershipconnection]

Oooh statistical noise disproves and invalidates math. This is just stupid. You see, there are formulas tied to descriptive and predictive models and theories, and then there are constant, unwavering laws. Sabremetrics has never claimed to be anything like gravity or the formation of covalent bonds. Rather, everything is based on data sets which describe and predict what SHOULD be, not what absolutely without exception WILL be. Some idiots think that anything that defies statistical predictions proves that stats have no merit. Yes, the mathophobes are right, these are games played by people and not computers, and that's exactly why they're wrong when they make stubborn and ignorant proclamations of the uselessness of stats.

 

Oh yeah, and clutch doesn't exist. It's noise. Look at those players over their careers and you will not find cluthiness.

[TuesdaysWithRillo]

I'm not trying to disprove sabermetrics here. I just thought I'd show an anomaly and see how you think the Mariners are beating the odds.

[TuesdaysWithRillo]

Oh yeah, and clutch doesn't exist. It's noise. Look at those players over their careers and you will not find cluthiness.

I think that is very arguable, but I don't want to open up a can of worms.

[Oldcity]

To me, Arizona is an even bigger anomaly. Their offense is flat-out bad.

[dadofandrew]

I do think there is some common sense in that teams that outperform where they should be based on runs scored have a very good bullpen, almost without exception.

 

All the rest though is cliched falsehoods. There have been 10 players in the history of the game who have hit substantially better in "close and late" situations, the last one was Mickey Tettleton...no, I don't think many of us called him clutch when he played.

[TuesdaysWithRillo]

Clutch is more than close/late situations. What about in the playoffs? ARod, Bonds, Maddux, and Glavine come to mind as guys that choke in the playoffs. Are their samples not big enough?

 

The Mariners could be a pesky team if they make the playoffs.

[homer]

"Clutch is more than close/late situations. What about in the playoffs? ARod, Bonds, Maddux, and Glavine come to mind as guys that choke in the playoffs. Are their samples not big enough? "

 

Frankly, yes. Anyone can go cold over the course of 50ABs. Sometimes it just happens in October.

 

And Maddux was pretty good in the playoffs - 3.34 ERA, 1.24 WHiP, in 194 IP. His career numbers are 3.10 and 1.14.

Same with Glavine: 3.42 and 1.27 in 218.1 innings. His career numbers are 3.49 and 1.3 - nearly identical.

 

The theory is that everyone comes back to their average with enough innings pitched or at bats. In the case of ARod he's had 132 total playoff ABs of which he's gone 11 for his last 60. I guarantee he's had similar stretches like that in his career. And if you look at the first 70 ABs ARod had in the playoffs he was 26 for 72 (.361) - I'd hardly call that choking. But given his career BA of .305 he was bound to regress eventually and he has. If he keeps getting playoff ABs I bet he trends upwards towards his career numbers.

[TuesdaysWithRillo]

Sometimes it just happens in October

 

When the pressure is on.

[homer]

When the pressure is on.

So what does it mean when they go cold in May? And is choking the only reason someone goes 1 for 12 in a playoff series?

[dadofandrew]

The playoffs are a tiny sample, even a 7 game series is only 30 AB's. Meaningless.

[danzig6767]

The Mariners have alot of high batting average hitters which generally will improve their chances of hitting at a high clip with RISP compared to teams who don't have alot of high batting average hitters.Compare Seattle to the Brewers,as a team Seattle hits .284 and we hit .263,it shouldn't be a surprise they do much better average wise with RISP.

 

As for their success,i've long felt that bullpen performance is somewhat underrated in baseball and their pen is top notch.Their rotation though outside of King Felix/Batista is so bad,i do think they will struggle to keep winning at this pace.

[Phenomenal Smith]

Was Reggie Jackson clutch?

 

I ask because everybody knows he killed the ball in the World Series, but how many of you know he hit only .227 in the LCS games that he played in?

[homer]

Reggie Jackson career post season numbers:

 

.278/.358/.527 in 281 ABs

 

Reggie Jackson career numbers:

 

.262/.356/.490

 

His Mr. October moniker was largely based on three series' over two years: 1977 World Series where he was 9/20 with 5 home runs and the 1978 ALCS (.462/.529/1.000) and 1978 WS (.391/.500/.696). This represented 56 ABs total out of 281 - I'm sure he had stretches like that in his career.

[Ennder]

Clutch is more than close/late situations. What about in the playoffs? ARod, Bonds, Maddux, and Glavine come to mind as guys that choke in the playoffs. Are their samples not big enough?

 

The Mariners could be a pesky team if they make the playoffs.

 

ARod pretty much proves the point of the sabermetic crowd. Before his last two post seasons he was extremely 'clutch' in the playoffs, the last two he has been horrible. So did he forget how to be clutch or something?

 

Generally speaking good bullpens win close games so if your bullpen is very good you tend to do better than your RS/RA suggests. There is also luck involved, Arizona is a great example as they have just been amazing in 1 run games this year 26-16 while being horrible in blowout games 12-21.

 

To use the Brewers as the counter example they are 19-14 in close games but 15-14 in blowouts which is pretty normal stats for a good team.

[whizkid]

What about in the playoffs? ARod, Bonds, Maddux, and Glavine come to mind as guys that choke in the playoffs.

 

Wouldn't you expect all players to do worse in the playoffs facing only the best teams, pitching staffs, lineups etc. Regular season stats are padded facing the bottom teams. But it does come down to a small playoff sample.

[Fear The Chorizo]

To me, Arizona is an even bigger anomaly. Their offense is flat-out bad.

That's the beauty of baseball - there are many different ways to win baseball games, and the team with the best lineup or best starting rotation isn't always guaranteed to win...many times teams that aren't overly great in any one facet of the game, but who are average to solid in every facet (starting pitching, bullpen, timely hitting/OBP, defense, baserunning, AND in-game managing) that end up having the most success over the course of a season. Teams that have any significant weaknesses get exposed - Milwaukee's biggest weakness has been pitching in the 6-7th inning, due to a severe lack of ability for their starters to get deep in games. For example, the Brewer bullpen has the arms to be very effective, but they've been severely taxed with the innings they've had to work.

 

The Dbacks have a very good, deep bullpen, and really a piecemeal rotation, headed by a true ace (Webb) that can get into the 8th and 9th innings basically every time he takes the mound. Their offense is sporadic at best, but they're winning while allowing their young talent to break into the bigs. They are pretty solid defensively, and their athleticism really shows up out in the field. They are a team that won't scare anyone come playoff time, and I think they'll have to really fight with the Padres and Dodgers to hang on to the division, but they have as much upside on their roster as the Brewers currently do - it should be fun the next few seasons watching the young players on the Brewers and Dbacks really come into their own.

 

As for sabermetrics, it's a great way to project and predict things, but like any other tool, it shouldn't be the end-all, be-all factor in building a ballclub.

[Phenomenal Smith]

A lot of people don't understand the purely random nature of small sample sizes. Here's an easy way to look at it. It you have a guy who is .280 hitter, your going to get all sorts of variations in performance over small sample sizes. This seems obvious, but here's an experiment to illustrate. I generate 30 random numbers in excel. If a number is 1-28, it's a hit, otherwise it's an out. I did this 3 times for a total of 90 numbers. In the first segment he "hit" .133, in the second he hit 333, and in the third he .300 for a total of .256.

[lukevan]

If a guy goes 30-for-50 in his first 50 ABs and then goes 0-for-50 he's still a .300 hitter. If every .300 hitter got EXACTLY 3 hits for every 10 consecutive ABs baseball would be pretty boring.

 

Go check out firejoemorgan.com if you want to read a dissection of the myth that is A-Rod's playoff career.