2019 Phillies Season Discussion (part 2)

[Taudawg]

Robertson will be 36 coming off TJS. The Phils might be better off considering the buy out as sunk cost.

[zambonir]

Probably but sometimes players come back strong from TJ rehab and he may not get more than 2 or 3 million guaranteed on open market. He may be amenable and even feels like he owes something to the Phillies if they want him.

[bkox]

Phillies catch a bit of a break. Tuesday was supposed to be Nola vs Sale in Boston. Sale has just been placed on the IL and will be seeing Dr James Andrews

[phillychuck]

I feel bad for Sale, one of the best pitchers in the game over the last several years.

[Austinfan]

Another data point counseling against giving long-term deals to pitches with significant mileage on their arms.

Red Sox signed him to a 5 year extension before the season, starting in 2020, 5yr, $145M, with a player opt out in 2022 - so if he bounces back in 2021 he can opt out, if not, Red Sox are on the hook.

Last year he started 30 games, including post-season, 173 IP. After throwing 673 IP the previous three years.

[zambonir]

Well, I imagine Sale is still better than whoever we might face, but might have been okay to catch him right now too.

I feel bad for Robertson!

[BostonRoss]

There are a lot of people where I live that feel bad about Sale.

I can sympathize. Sign a guy to a five year extension, and he gets hurt before the extension starts.

[BostonRoss]

Most likely it will be Brian Johnson, who has a 6.45 ERA and a WHIP over 2. There’s a possibility they could activate Price, or move Cashner back to the rotation, where he’s been awful.

[allentown]

Sign a guy to a five year extension, and he gets hurt before the extension starts.

Yes, that does bring back a familiar ache, doesn't it?

[VoteForKalas]

Since 2003, only 1 team has finished the regular season with a negative run differential and made the playoffs. That was the 2007 D-Backs who won the NL West and were -20.*

In that same time frame, the lowest wild card finisher in run differential was the 2012 Orioles at +7*

Phillies are currently -18 and 2 GB of the second wild card spot - as Captain Obvious would say, unless this trend changes very soon the odds of the Phillies making the playoffs is slim.

*Source: The Athletic Aug 17th Phillies article

[zambonir]

I feel like the run differential will be largely self-fulfilling. If they actually pass Chicago or Washington they might get back over zero (because I think they'll have to win more 8-4 games to do that). Going by run differential alone there isn't really even much of wild card race, but the Cubs, Brewers and Cardinals play each other a lot. I don't think we are much worse than any of those teams but it seems unlikely we'll be close enough to the Nats to catch them in late September.

[VoteForKalas]

Phillies are playing over their heads - their Pythagorean W/L record based on run differential is 60-64

I agree on paper it does not seem we are worse than the Cubs, Brewers or Cards, but there must be some set of reasons why our run differential is. I wonder if it is more on the run production or run prevention side.

[andyb]

Part of it is blowouts. We have lost 3 games by 14 runs. Biggest wins have been by 11, 10, and 10. This is a little bit of a sign of a thin bullpen, but one could argue we get a minus 10 net on games where we have stopped trying at some point.

Though the larger point I guess is still that if we do make the playoffs we'll probably be at least +20 the rest of the way. Clearly this is not a great team, but if the offensive performers had met expectations we'd probably be a small net positive even with the uncertain bullpen and starting staff.

[zambonir]

I'm generally a fan of Pythagorean but I would think +/- 3 or 4 wins is not significant enough to say they are above their heads (compared to say, the D-Backs, or the Reds, who are underperforming).

Put another way - our X W-L is 60-64 and the Mets X W-L at 64-60 (also their actual record) and I'm not sure there's actually a serious qualitative difference between those two records (or 62-62).

My guess is, for all the focus on the offense and Mallee, that pitching has 100% been the difference in the times when this team has been good and when it has not. I mean, maybe the stats would say otherwise, but crudely it's safe to say that when 80% of your rotation and 75-88% of your bullpen has cratered, that's a bigger problem than Rhys Hoskins' BA with RISP. We were leading the division and doing fine back when Arrieta was still decent and Eflin was a potential All-Star, though the offense also carried Nola.

It's just that the offense living up to or exceeding its potential is their only real path to wins.

(I mean, having Jason Vargas pitch great every night and having Eflin return to his April form would also do the trick, but seems less likely. I think the bullpen has been pretty solid lately but they still can't protect every tie or one-run lead, whether it's 3-2 or 6-5.)

[VoteForKalas]

I'm generally a fan of Pythagorean but I would think +/- 3 or 4 wins is not significant enough to say they are above their heads (compared to say, the D-Backs, or the Reds, who are underperforming).

Put another way - our X W-L is 60-64 and the Mets X W-L at 64-60 (also their actual record) and I'm not sure there's actually a serious qualitative difference between those two records (or 62-62).

It may not be definitively significant, but hopefully we can agree that 64-60 is better than 60-64

I agree it appears the pitching ups and downs have set most of the tone and that the offense likely will need to over-perform to get us to the playoffs.

[allentown]

I'm generally not a fan at this stage of the season. The Pythagorean predicts a result for which we already have an exact measurement. If random luck has influenced the win total after 124 games, I see no reason to assume that the Pythagorean gives a meaningful hint of what is most likely to happen in the final 38 games. Statistically, 124 isn't a huge sample size, but 38 is a really SSS, so to expect some drive toward W-L record converging to/toward the Pythagorean prediction by end of season seems problematic to me. I think current, recent performance does have some validity, since it incorporates injuries, acquisitions at trade deadline, rookies early season success being tempered as the league probes for weaknesses. If asked to give a prediction for end-of-season record, I'd bet more on just assuming that the last 38 games would have either the same W-L as the most recent 38 games (possibly with some adjustment for strength of schedule) or the same percentage of wins as the team has had since the trade deadline (again with a possible adjustment for strength of scheduled opponents for remaining games).

While it is statistically certain that if I put 500 red balls and 500 black balls in a jar and pick them out one by one, having pulled 300 black balls on my first 500 draws means I absolutely must pull only 200 black balls on my final 500 balls, the Pythagorean is not working from such a tightly-bounded universe with a final conversion which MUST be converged upon. As mentioned above, blow-outs, one direction or the other, have an out-sized influence on the Pythagorean. I think managerial choice has at least some impact on the magnitude of the average blow-out.

[David_Criss]

Quick question on Franco. Noticed today he played 8/8 and 8/9 and then not again until last night. Was he hurt? I am guessing so.

[phillychuck]

@allentown

Regression to the mean is real, even in small samples. If you flipped a coin 124 times you'd expect 62 heads...if you got 68 heads, the 38 games remaining would tend to split 19/19, moving your "head" percentage closer to the actual skill level of the coin, 50%. The runs stat shows us to be closer to a .500 team or slightly below, meaning we should expect to finish around 18/20 over the last 38 (not adjusting for schedule strength). So we should regress to the mean if the run differential is a true measure of our skill level. That can be argued...

[Patchmk88]

Yea he was on the injured list.

[zambonir]

Yeah I don't know if X W-L is fully projectable but it's safe to say that if the Phillies finish with a negative run differential they also probably won't have significantly outperformed it, and if they don't outperform it they will have missed the playoffs, just as they would today. It seems to me that for all the nits we pick about blowouts (or close games) the formula is rarely an incorrect snapshot of a team once the season is over.. Honestly, not that I have blind faith in the statheads, but if people have made this argument for as long as the equation has existed; if one of the sites felt the formula needed to be weighted or adjusted for blowouts to be more descriptive or predictive they would have done it (and some people have, on their own, and perhaps it tells a better story at any given point in a season.

Anyway, on June 15 we had a nice positive run differential (+20) and were 39-31 with an X W-L of 37-33. That team had McCutchen, a viable Arrieta, a dominant Zach Eflin, Scott Kingery on the bench (even if the guys "ahead" of him weren't contributing much) and a few more bullpen arms (though not many). The skill level of the team that played the next six weeks was not the same, nor is it now. It's been a .500 team or slightly worse (sometimes not just slightly) for quite a while, enough to swing the differential from +20 to -18. Vargas and the relief dumpster dives and the hot bats help but as we quickly saw, this team is still capable of following a sweep with a 1-2 series loss. But so are the Cubs and Cards and Brewers.

Things all tend to even out over the 162. If the Mets really are four games better than us now we'll find out soon enough. And the Nats may even be better than the Braves now (including by X W-L, slightly) but might not have enough time to catch them.