Poor Kevin Cash

[Bruce Berenyi]

this is not ann exact science he should have kept his eyes open in the fifth inning when snell was dominant ,, at least let the next inning proceed before intervening

[PuffyPig]

In addition to this, just look at the situation. Snell is by far their best pitcher, he's coming off a short season where he pitched 50 innings. In a regular year, it's the equivalent of the middle of June, mid-season form. He'd thrown 73 pitches. This was literally the last time he was going to pitch this season, so there was no concern that pushing him would impact his next start. It was win or go home. Analytics or no, he goes back out there until something tells me he needs to come out. If somebody gets on base, his velocity drops, his command goes wonky...but none of that was happening.

The Rays didn't score enough runs, so it probably didn't really matter in the end, but removing Snell was one of the worst decisions I've ever seen a Manager make.

His velocity was dropping though.

[M2]

The Rays didn't score enough runs, so it probably didn't really matter in the end, but removing Snell was one of the worst decisions I've ever seen a Manager make.

It 100% didn't matter. The Dodgers were going to get to whoever Cash put out on the mound. Like Arthur Jarrett in Monty Python's The Meaning of Life, Cash was just choosing the method of his own execution.

[Boston Red]

Personally, I think the Dodgers scoring there hurt the Rays offense as well. You're a lot more comfortable at the plate when your team is in the lead.

[Chip R]

I was looking on Baseball Reference yesterday and noticed that Snell had never pitched over 5.2 innings this season. Whether you agree with Cash pulling him that early, Snell was getting into uncharted territory in Game 6. Perhaps he could have got another inning or so out of him on Tuesday but I have to believe that was one thing Cash took into account.

[M2]

His velocity was dropping though.

It's kind of weird how so many people arguing the manager should have been paying more attention to what Snell was doing in that game aren't paying attention to the velo drop or that Austin Barnes just squared him up.

[PuffyPig]

Look, we all get that the decision didn't work out well, but we have no idea how leaving him in would have worked either.

Snell had been injured during the season and had not logged many innings. He certainly wasn't a guy like Bauer.

The decision to take out Snell was defensible I think.

I think the choice of bringing in Anderson was the one that should be questioned more so.

[757690]

If analytics suggests a move is successful 60% of the time, it’s a good move over a full season. It’s also a good move in a short series even though it will fail 40% of the time.


Sent from my iPad using Tapatalk

You’re missing the point.

The percentages don’t have time to work themselves out in the playoffs. It’s do or die. The percentages only work if you can make that decision 100 times. Make it 3 times or once, and the percentages are less important because they won’t have time to work out. What matters is the outcome of that very decision, not the outcome that decision 100 times over the course of a season.

In poker, if you are at the final table, final hand, and you need the next card turned over to be a heart, it doesn’t matter what the odds are, what matters is what the suite of that next card is. The odds are less important than what the next card actually is. You can’t think, “I may get this one wrong, but over the long haul, the odds will work themselves out.”

The smaller the sample size, the less important the percentages are, and the more important that you get this very call right.

[PuffyPig]

You’re missing the point.

The percentages don’t have time to work themselves out in the playoffs. It’s do or die. The percentages only work if you can make that decision 100 times. Make it 3 times or once, and the percentages are less important because they won’t have time to work out. What matters is the outcome of that very decision, not the outcome that decision 100 times over the course of a season.

In poker, if you are at the final table, final hand, and you need the next card turned over to be a heart, it doesn’t matter what the odds are, what matters is what the suite of that next card is. The odds are less important than what the next card actually is.

The smaller the sample size, the less important the percentages are, and the more important that you get this very call right.

If you ignore the percentages in a smaller sample size, then you should always ignore them. The fact they don't have time to "average out" because it's one game doesn't mean you should ignore the odds which are in your favor.

Are you suggesting that you ignore odds in smaller sample sizes? I would suggest that you are more likely to get the call right by going by the percentages.

[M2]

You’re missing the point.

The percentages don’t have time to work themselves out in the playoffs. It’s do or die. The percentages only work if you can make that decision 100 times. Make it 3 times or once, and the percentages are less important because they won’t have time to work out. What matters is the outcome of that very decision, not the outcome that decision 100 times over the course of a season.

In poker, if you are at the final table, final hand, and you need the next card turned over to be a heart, it doesn’t matter what the odds are, what matters is what the suite of that next card is. The odds are less important than what the next card actually is.

The smaller the sample size, the less important the percentages are, and the more important that you get this very call right.

The next card in this case was Blake Snell throwing a diminished fastball to Mookie Betts and then possibly Corey Seager. "Maybe he'd have gotten away with it" isn't a particularly compelling argument. Snell gave what he had and his tank was just about empty. As has been noted, he's not a guy who pitches deep into games.

[Patrick Bateman]

Baseball reference has Snell's splits by pitch count (career). Not surprisingly, he is historically far less effective after 75 pitches, which he was right on the cusp on, with signs of velocity decline. There's just no real reason to believe he was going to be dominant like he was the first 5.2 innings.

[Coopdaddy67]

Look, we all get that the decision didn't work out well, but we have no idea how leaving him in would have worked either.

Absolutely, but if I'm going to lose the game, I'm rolling with the former Cy Young winner that's cruising through 5 innings.

Meanwhile, Anderson gave up a run in all 3 games he appeared during the World Series. He gave up a run in his last 7 playoff appearances, 8 out of 10 appearances total.

Not surprisingly, he is historically far less effective after 75 pitches, which he was right on the cusp on, with signs of velocity decline.

I think the velocity decline argument is quite the stretch, especially when you're looking at only a few pitches. Based on previous starts this year and during his Cy Young season, he bounces around all game long. Even the charts provided in the fangraphs article show him all over the place velocity-wise, which is by design. You can see the same thing in his other starts this year and during his Cy Young season.

[Patrick Bateman]

If you ignore the percentages in a smaller sample size, then you should always ignore them. The fact they don't have time to "average out" because it's one game doesn't mean you should ignore the odds which are in your favor.

Are you suggesting that you ignore odds in smaller sample sizes? I would suggest that you are more likely to get the call right by going by the percentages.

This 100%.

Just because you hit the casino and hit a few roulette spins doesn't mean it was favourable or you should keep doing it as an investment because you fluked out once with your money on the line.

You play the odds, based on the best evidence available, and in a small sample size, there are no guarantees. You simply try and align your odds as favourably as possible. In this case, the evidence would be a combination of real time velocity, historic pitch count and times through the lineup statistics, opponent analysis, etc.

It was a tough decision, but there is compelling evidence that supported the removal of Snell for an elite, fresh pitcher. Not sure Anderson was the guy in this particular case, but lets not act like getting torched by Betts wasn't a potential outcome for anyone, and should judge the end result solely because an elite hitter succeeded at the plate.

[JFLegal]

everyone saying something to the effect of "well, he was just going with the analytics and that's what got them there" needs to watch the jomboy video that was linked in the original post. not only did pulling snell go against common sense, it also went against the analytics.

and the fact that snell had not pitched more than 5.2 innings in a game this season is irrelevant. we see things happen all the time in the playoffs and especially the world series from pitchers that we do not see in the regular season. when the dbacks won the title, randy johnson was used as a reliever on 1 day rest (i'm going off memory so forgive me if I don't have that exactly correct). he never did that during the regular season, but it was the best way to win. i believe the giants did the same thing with madison bumgarner.

the point is that snell not doing something during the regular season should have nothing to do with the situation he was in there with the world series on the line. he had only thrown 73 pitches and was near perfect. mookie betts gets WORSE the more he sees the same pitcher in a game. mookie betts does not fare well against left-handed pitchers like snell. mookie betts DOES fare well against right-handers who rely on their fastball. what does kevin cash do? take snell out and put anderson in. kevin cash didn't really study the analytics there. he just came up with his own gut feeling and thought he was going with what the analytics said. jomboy's video points this out in very clear terms for anyone who might be confused.

[757690]

If you ignore the percentages in a smaller sample size, then you should always ignore them. The fact they don't have time to "average out" because it's one game doesn't mean you should ignore the odds which are in your favor.

Are you suggesting that you ignore odds in smaller sample sizes? I would suggest that you are more likely to get the call right by going by the percentages.

I didn’t say ignore them, just that the are less important.

Over a full season, making the decision based solely on the percentages makes sense.

In one game, other factors become more important and should be taken into account when making the decision.

My issue is with the dogmatic adherence to the plan based 100% on percentages. It works well in a 162 game season. In a 7 game series, or in an elimination game, it doesn’t work as well. It’s why the A’s don’t do that well in the post season. Heck, even Billy Beane has admitted as much.