Kimmi

I agree that ERA is not the worst stat ever. That would go to Wins, Saves, and Holds, not necessarily in that order. However, judging a pitcher by ERA alone would be a huge, huge mistake. I don't think anyone here does that, but there are still plenty of people who do.

Fair enough. I do agree that there are pitchers who can be very effective with low K rates. I think the point that both of us agree on is to use as many tools as possible to get the best overall assessment.

Well, before we acquired Sale, I said that my biggest concern for our team was the lack of starting pitching depth. If ERod's injury is significant, our depth takes a huge hit before the season even starts. Where we once had 8 viable starting pitchers were debating about who should start over whom, we could possibly be down to 5 viable starters and be in serious trouble if we have another injury.

Absolutely. The smaller the sample size, the shorter the series, the closer the score in the game, the more randomness trumps skill.

I think that Swihart could be an above average catcher defensively if given the proper time to work on his development behind the plate. Vazquez has been touted as being defensively elite. Therefore, any comparisons of Swihart to Vazquez defensively have left Swihart falling short, which is unfair to him. I hope that Vazquez' injury was the reason for his step backwards last year, and that he can reclaim his 'elite' status. If I had to bet, I would bet that Swihart will not reach the same level defensively as Vazquez.

I see it as the reverse. Some of you are trying to assign an explanation to why everything happens, be it tangible (mechanical adjustment) or intangible (mental sharpness). More often than you think, there really is no explanation other than just pure luck.

Not all changes in performance are due to randomness, especially when you're talking about year to year performance. No one has suggested that. The smaller the sample size for the improvement or the slump, the larger the role that randomness likely plays. I acknowledge that factors besides randomness could be the reason for a hot or cold streak, especially as those streaks become longer. However, I do believe that randomness plays a larger role in baseball than most people are willing to give it credit for. I am also positive that while randomness was not the driving force in the difference between the numbers you listed above, it did play a part.

I read that earlier and did a mental eye roll to myself. I understand the team wanting to get under the salary cap, but I was afraid the trade might come back to haunt us. Not that it has yet. I certainly hope ERod's knee injury is nothing serious.

They're not going to accept something is true just because people say it is or because it seems so right. Doing that would be nonsense. The best thing about the stat geeks is that they don't rest on their laurels. They will continue to test these things, and as new data and new technology become available, they will retest with improved methodology. They also double check and retest each other's work. They may even prove each other or themselves wrong, and that's okay with them. As I said before, it's much better than to insist the world is flat, because that's what you've always known.

No, it's not a description of the results. Randomness explains why things happen. I don't know about basketball, but I can give you example after example in baseball.

If you are directing that comment at me, you're missing the boat. I had an opinion based on what I thought I saw. The stats didn't agree with me. I was wrong and I changed my opinion. You have an opinion based on what you think you see. The stats don't agree with you so therefore the stats must be wrong.

I have to agree with Notin in his past two posts. He is spot on. The players that we think of as 'clutch' do not become better players in the postseason. They are simply good players all the time. Papi does well in the postseason because he's simply a good hitter.

The evidence I have read is not anecdotal. It is statistical evidence. Listing a bunch of examples is anecdotal evidence. It does not prove or disprove anything. It is widely accepted in the sabermetric community that Ks for pitchers are very good and that SIERA, FIP, etc. are better predictors of future ERA than ERA or ERA- are. I am pretty sure that I've provided links to you before showing the statistical evidence. Fangraphs has a very cool tool that allows you to find the correlation between any two pitching stats. (There is a separate tool for hitting stats for anyone interested.) Here is how different stats correlate to ERA- the following year: K% -.361 K/9 -.341 ERA- .301 WHIP .260 wOBA against (They don't have OPS against, but wOBA is better anyway) .301 SIERA .390 FIP .331 xFIP .341 Surprisingly, simple K% does a better job in this study than FIP or xFIP does. I've seen other studies where FIP and xFIP are better, but that was comparing to future ERA, not ERA-. But you can see that any of the DIPS stats, which rely heavily on Ks, do better than the 3 stats that you mentioned. I'm not saying that ERA-, WHIP, or wOBA are bad stats. They aren't. Here is the link if you want to play around with it yourself. http://www.fangraphs.com/blogs/tool-basically-every-pitching-stat-correlation/

Absolutely. If the stats tell me something different than what I have believed from the eye test, I'm going with the stats. The eye test involves all human element, which we know is subject to things like bias. I used to think that Pedroia was a great base runner. That's what I saw when I watched the games. The announcers used to confirm what I believed. Or maybe the announcers saying that Pedroia was a great base runner influenced my opinion into 'seeing' that? Either way, when the stats told me that Pedroia was not a great base runner, I had to rethink my opinion on that and admit that I was wrong.

Randomness.

No one has ever said otherwise.

Excellent post Hugh. I agree.

The whole thing about the hot hand fallacy is the belief that having a hot hand has some predictive value. It doesn't. Just like momentum exists but it has no predictive value, hot streaks exist, but they have no predictive value. For a manager to play a weaker overall hitter who is 7 out of his last 10 over a stronger overall hitter who is 1 for his last 10 would be a mistake, unless there is an identifiable reason why the 1 for 10 player is in a mini slump.

You are right. However, they often can tell you why something happened, or at the very least, they can eliminate reasons for why something happened. When dealing with the human element, there are always going to be imperfections in the stats, and there will always be things that simply can't be measured. There is not a single stat geek who doesn't acknowledge this. However, as I've said before, determining why things happen is pretty much their reason for existence. They do a pretty doggone good job of it. IMO, it's far better than simply making a statement based on observation or perception and accepting it as true because we believe it to be true.

I love the work that Bill James and other stat geeks do. Contrary to the way many people feel about sabermetrics, I think it makes the game of baseball far more interesting and enjoyable. I have found that while it's not always the case, people who criticize sabermetrics usually do so for one of two reasons: 1. They don't understand it. 2. The particular stat or research does not support what they have always believed to be true. I don't fault anybody for either of the two reasons. Stats are difficult to understand or are simply not a topic of interest for some, and the human element aspect of baseball is a strong and very real one. That doesn't make the research wrong though.

You can disagree with it all you want, but it's not nonsense.

I know I'm not going to convince you all on this. As I posted earlier, I'm having a hard time convincing myself. I have had those moments where I've been 'in the zone' myself. Athletes are firm believers in this concept. Bill James had an interesting article on the topic. He relates it to his pool shooting experience. He is 100% convinced that not all of the variance in streaking is random, but he goes on to say that 95 to 99% of it is due to randomness. Otherwise the causal factors would show up in the data. Here's a link if you are so inclined to read the article. If you do, also read the comments (from the bottom up) and click on the related thread link that Tangotiger provided in the comments and read that exchange. Good stuff. http://www.billjamesonline.com/hot_hand_question/

Vazquez ranked 22nd out of 114 catchers in terms of extra strike calls per game, and 21st out of 114 in pitch framing runs above average, per StatCorner. At Baseball Prospectus, Vazquez ranks 12th overall defensively out of 114 catchers. He was 17th in framing runs (+7), tied for 21st in blocking runs (+.9), tied for 38th in throwing runs (at exactly 0.0), and 11th in fielding runs (+11). I'd say that's very good, but still a drop off from his 2014 numbers. To be fair, Fangraphs DRS has him at -1.

I agree. Not everyone does though.

That is a valid point. LOL about JoeBrady Country. Joe remains one of my favorite posters. You are absolutely right about him always citing K/BB. There is a reason for that. It's a very simple stat, yet very effective.