Roadmap to Contention 2020

[SteveA]

16 hours ago, MurphDogg said:

What I find most interesting here is that the Orioles at 9-7 (possibly including their being heavy favorites in the suspended game against the Nats tomorrow) have MUCH better odds than the 8-4 Marlins (30.1 percent) and the 9-7 Tigers (13.0 percent).

That just doesn't seem right.   We are still in 3rd place, so we are in position for the 7 seed (#1 wild card) and just 1 game ahead of both the Indians and White Sox. 

So all it would take would be a 1.5 game slip in the standings to Chicago and Cleveland and we would not be sitting in a playoff spot.

It's hard to see how that translates to a 52% playoff probability.

[Filmstudy]

 

1 hour ago, DrungoHazewood said:

A true talent estimate of the Orioles in normal circumstances would be something like 50 wins.  Certainly no more than 60, since they won 54 last year and subtracted talent.

It's not terribly unlikely for a .325 team to go 10-7 at some point in the schedule.  Last year they had some periods of 15-ish games at .500, and at least one run of 7-3.  I think there's about a 25% chance of a .325 team going 10-7 in any 17 game stretch in a 162-game schedule.

I think they're going to need to be around .500 to sneak into the playoffs this year.  That means they're going to need to go 20-23 the rest of the way.  So we need to know the odds of a .325-ish team going 20-23 or better.  You can estimate that with a binomial probability calculation, and the odds of a .325 team winning 20 or more of 43 games is 3.8%.

But I'm guessing optimistic Oriole fans think the O's are not a .325 team, but in fact a .400 team or better.  To get a 50% chance of 20 or more wins you need to assume that they're now a .455 team, or the equivalent of a 74-win team in a 162-game schedule.  To me that seems a little wacky given last year's results and the current roster composition.  But I guess we'll see.

I agree that 0% odds of winning the Series seems wrong.  An average playoff team would have a 6.25% chance (1-in-16).  If you assume the Orioles are a .455 team playing .550 teams their odds wouldn't be good but they'd be better than 0%.  Using that same binomial distribution, a .400 team wins a 3-of-5 series something like 30% of the time, and a seven-game series just a bit less.  So winning three straight would be about (.3)^3, or about 3%.  So 50% chance of getting to a playoff where you win it all 3% of the time, so I'd peg the O's optimistic odds at a trophy at about 1.5%.

Good to see the math approach.

First step in determining if the model is reasonable is to determine if .325 can be a reasonable estimate for a team that goes 10-7 in the first 17 games.  The chance of a .325 team going 10-7 or better in the first 17 games of a season is only 2.3%, so I'd argue we've already seen a level of statistical significance that the Orioles do not truly have a team with .325 talent.  Said otherwise, the hypothesis that the Orioles are a .325 winning percentage team can already be rejected at the 90%, 95%, or even 97.5% confidence level.

Taking another approach, you have the Orioles jumping from a .325 expected win probability at season's beginning to a .450 winning percentage as a playoff team.  That doesn't seem unreasonable, but after 17 games of a 60-game season, shouldn't the Orioles have already "credibility graded" towards whatever they will be entering the playoffs from .325?  It's not an uncommon approach in actuarial work to have a standard you begin with (.325 is probably reasonable) and another based on actual experience to allow such grading (10/17 might be reasonable).  If you use something like that you'd estimate their current win probability at .325 X (43/60) + .588 X (17/60), which makes the Orioles a .400 (.3995) expected win probability team from this point forward.  That feels like a more reasonable number based on what we've seen so far from this team.  If I plug that back into the binomial theorm, I get:

30+ wins: 23.6%

29 wins: 10.4%

28 wins: 11.9%

Based on projections I'm seeing, the Orioles will probably be in with either 28 or 29 wins, but let's say it's 30, since the September Effect will split haves and have nots after the trade deadline.

As I understand the playoffs, they would then face (probably) the 1 or 2 seed in a 3-game wild card series.  In a way that's an ideal format for a bad team (toughest opponent in shortest series).  Then 7 game series the rest of the way.

Edited August 14, 2020 by Filmstudy

[Can_of_corn]

Second round is best of five.

[MurphDogg]

33 minutes ago, SteveA said:

That just doesn't seem right.   We are still in 3rd place, so we are in position for the 7 seed (#1 wild card) and just 1 game ahead of both the Indians and White Sox. 

So all it would take would be a 1.5 game slip in the standings to Chicago and Cleveland and we would not be sitting in a playoff spot.

It's hard to see how that translates to a 52% playoff probability.

No, they think that both Chicago or Cleveland will make it because one will be second in the AL Central and the other would still be the #7/8 seed and wouldn't deprive us of a playoff spot.

The reason our odds are so good is because they really seem to hate the Tigers and the Rangers. While we lead the Tigers by only half a game (and they are ACTUALLY in second in the AL Central), their odds are 8.7% while ours are 60.6%. Meanwhile the Rangers are currently in second in the AL West in playoff position and we lead them by two games but BR has them at 19.4%, while the third place in the AL West Astros are at 93.3%.

The AL team that has the 9th best odds of making the playoffs per Baseball Reference is the Blue Jays at 32.5%, and we have 3 games on them.

[Bubble Buddy]

54 minutes ago, DrungoHazewood said:

A true talent estimate of the Orioles in normal circumstances would be something like 50 wins.  Certainly no more than 60, since they won 54 last year and subtracted talent.

It's not terribly unlikely for a .325 team to go 10-7 at some point in the schedule.  Last year they had some periods of 15-ish games at .500, and at least one run of 7-3.  I think there's about a 25% chance of a .325 team going 10-7 in any 17 game stretch in a 162-game schedule.

I think they're going to need to be around .500 to sneak into the playoffs this year.  That means they're going to need to go 20-23 the rest of the way.  So we need to know the odds of a .325-ish team going 20-23 or better.  You can estimate that with a binomial probability calculation, and the odds of a .325 team winning 20 or more of 43 games is 3.8%.

But I'm guessing optimistic Oriole fans think the O's are not a .325 team, but in fact a .400 team or better.  To get a 50% chance of 20 or more wins you need to assume that they're now a .455 team, or the equivalent of a 74-win team in a 162-game schedule.  To me that seems a little wacky given last year's results and the current roster composition.  But I guess we'll see.

I agree that 0% odds of winning the Series seems wrong.  An average playoff team would have a 6.25% chance (1-in-16).  If you assume the Orioles are a .455 team playing .550 teams their odds wouldn't be good but they'd be better than 0%.  Using that same binomial distribution, a .400 team wins a 3-of-5 series something like 30% of the time, and a seven-game series just a bit less.  So winning three straight would be about (.3)^3, or about 3%.  So 50% chance of getting to a playoff where you win it all 3% of the time, so I'd peg the O's optimistic odds at a trophy at about 1.5%.

Here's my question when taking last year and using it as a baseline assumption for this year: At what point are we able to factor in the improvements of our young players and consider that they may be a .400 team or better? I know you said I guess we'll see, and reality is, that's all we can really do at this point. But I'm curious when you will consider it reasonable that they could play .455 ball the rest of the way? 

 

I am as surprised as anyone, having referred to a lot of these guys as scrapheap guys. None of them really have prospect pedigree, or at least if they did, they went through some serious rough patches. And I agree that this is a SSS. Fortunately for us, the whole season is a SSS. 

I do think there is the slight possibility offensively that we've hit the lottery and have 5-7 unheralded (or slow developing) young players that have grown at once (Ruiz, Severino, Sisco, Scott, Castro, Nunez, Santander, Alberto). The good news with Severino, Alberto, Santander and Nunez is they gave us a little bit of a sampling from last year. If they've made incremental improvements on who they were last year, they are pretty good role players. 

And then as far as starters, it helps having Cobb back and hopefully Means can stay healthy (or really both of them!). It could almost be like we swapped Cobb for Cashner. The question then becomes, can Means produce similar to last year down the stretch and can we get someone duplicate Bundy's production and not have the 4/5 roles be a complete disaster. 

As far as the bullpen, right now our team ERA is 4.19, a huge improvement on last year's 5.79. Plus I think the peripherals point to that ERA being sustainable and not too fluky. I like that we've taken Givens out of the closer role and I like what I've seen from Tanner Scott. And hopefully we can get Hunter Harvey back. 

I'm not savvy enough to get the advance stats on the defense and would be curious about that. It seems like we're avoiding the costly mistakes (when Nunez is DH), but I know there's a lot more to it than that. 

But, all that to be said, if the goal is to play .455 baseball from here on out (possibly slightly less if we can sneak that Nats win), that doesn't seem crazy anymore in this 60 game season with the start we've had. 

[Aristotelian]

1 hour ago, DrungoHazewood said:

A true talent estimate of the Orioles in normal circumstances would be something like 50 wins.  Certainly no more than 60, since they won 54 last year and subtracted talent.

It's not terribly unlikely for a .325 team to go 10-7 at some point in the schedule.  Last year they had some periods of 15-ish games at .500, and at least one run of 7-3.  I think there's about a 25% chance of a .325 team going 10-7 in any 17 game stretch in a 162-game schedule.

I think they're going to need to be around .500 to sneak into the playoffs this year.  That means they're going to need to go 20-23 the rest of the way.  So we need to know the odds of a .325-ish team going 20-23 or better.  You can estimate that with a binomial probability calculation, and the odds of a .325 team winning 20 or more of 43 games is 3.8%.

But I'm guessing optimistic Oriole fans think the O's are not a .325 team, but in fact a .400 team or better.  To get a 50% chance of 20 or more wins you need to assume that they're now a .455 team, or the equivalent of a 74-win team in a 162-game schedule.  To me that seems a little wacky given last year's results and the current roster composition.  But I guess we'll see.

I agree that 0% odds of winning the Series seems wrong.  An average playoff team would have a 6.25% chance (1-in-16).  If you assume the Orioles are a .455 team playing .550 teams their odds wouldn't be good but they'd be better than 0%.  Using that same binomial distribution, a .400 team wins a 3-of-5 series something like 30% of the time, and a seven-game series just a bit less.  So winning three straight would be about (.3)^3, or about 3%.  So 50% chance of getting to a playoff where you win it all 3% of the time, so I'd peg the O's optimistic odds at a trophy at about 1.5%.

I think this is a little pessimistic. Looking at it the other way, what are the odds of a team that starts 10-7 actually being a .325 team the rest of the way? This is not a random sample of games at some cherry picked point in the season but it's what we have to go on. We lost Villar but added Iglesias and increase Alberto's role. We lost Mancini, but add Hays and more AB's for Nunez and Smith, plus Santander is a defensive upgrade. Cobb appears healthy. There are some guys playing over their heads but I also see Hays underperforming, Davis closer and closer to losing his job permanently, and Means due to come back. In an evaluation year we had a ton of innings from guys like Straily, Ynoa, Yacabonis, Blach etc who are all gone. All of which is to say I think there is some reason to think this could be an improved team despite losing Mancini and Villar. Of course, I am not predicting that they will end up .600 but seeing what we've seen I think it's possible to say this is an improved team over last year.

[DrungoHazewood]

1 hour ago, Filmstudy said:

First step in determining if the model is reasonable is to determine if .325 can be a reasonable estimate for a team that goes 10-7 in the first 17 games.  The chance of a .325 team going 10-7 or better in the first 17 games of a season is only 2.3%, so I'd argue we've already seen a level of statistical significance that the Orioles do not truly have a team with .325 talent.  Said otherwise, the hypothesis that the Orioles are a .325 winning percentage team can already be rejected at the 90%, 95%, or even 97.5% confidence level.

A little real life data to see how it compares to the probabilities:

There have been 16 MLB teams who finished with a winning percentage between .320 and .329.  Over their first 17 games they ranged from 3-14 to 8-9.  Overall they were 88-167, good for a .345 winning percentage.  So nobody who finished with a winning percentage in the .320s ever started as well as 10-7.

There have been 20 teams that finished between .330 and .339.  They ranged from 10-7 to 0-17 (that being the 1988 Orioles).  The 1908 NY Highlanders, soon to be Yanks, went 10-7 on their way to 51-103.  Two other teams were over .500 in their first 17 games on their way to 100+ losses, I believe those were the '27 Red Sox and '54 A's.  All of these .330 teams totaled 110 wins and 230 losses, good for a .324 percentage.

So it is unlikely but not unprecedented for a .330-ish team to start with a winning record over their first 17 games.

[DrungoHazewood]

One more thing, the 1998-2011 Orioles never had a single winning record at the end of the year.  Over that entire 13 year period I'm guessing they won about 43% of their games.

During that period they started the year 11-6 twice, and 10-7 five consecutive years (from 2004-2008).  I might grudgingly concede that this team is as good as the Kevin Millar, Danys Baez era Orioles. But I'm not in love with the concept.

[Filmstudy]

8 hours ago, DrungoHazewood said:

A little real life data to see how it compares to the probabilities:

There have been 16 MLB teams who finished with a winning percentage between .320 and .329.  Over their first 17 games they ranged from 3-14 to 8-9.  Overall they were 88-167, good for a .345 winning percentage.  So nobody who finished with a winning percentage in the .320s ever started as well as 10-7.

There have been 20 teams that finished between .330 and .339.  They ranged from 10-7 to 0-17 (that being the 1988 Orioles).  The 1908 NY Highlanders, soon to be Yanks, went 10-7 on their way to 51-103.  Two other teams were over .500 in their first 17 games on their way to 100+ losses, I believe those were the '27 Red Sox and '54 A's.  All of these .330 teams totaled 110 wins and 230 losses, good for a .324 percentage.

So it is unlikely but not unprecedented for a .330-ish team to start with a winning record over their first 17 games.

 

[Aristotelian]

9 hours ago, MurphDogg said:

No, they think that both Chicago or Cleveland will make it because one will be second in the AL Central and the other would still be the #7/8 seed and wouldn't deprive us of a playoff spot.

The reason our odds are so good is because they really seem to hate the Tigers and the Rangers. While we lead the Tigers by only half a game (and they are ACTUALLY in second in the AL Central), their odds are 8.7% while ours are 60.6%. Meanwhile the Rangers are currently in second in the AL West in playoff position and we lead them by two games but BR has them at 19.4%, while the third place in the AL West Astros are at 93.3%.

The AL team that has the 9th best odds of making the playoffs per Baseball Reference is the Blue Jays at 32.5%, and we have 3 games on them.

That is crazy. We might actually do this.

Will be fun if it really is us vs Blue Jay's. Could finally get payback for '89!!!

[Mr. Chewbacca Jr.]

9 hours ago, MurphDogg said:

No, they think that both Chicago or Cleveland will make it because one will be second in the AL Central and the other would still be the #7/8 seed and wouldn't deprive us of a playoff spot.

The reason our odds are so good is because they really seem to hate the Tigers and the Rangers. While we lead the Tigers by only half a game (and they are ACTUALLY in second in the AL Central), their odds are 8.7% while ours are 60.6%. Meanwhile the Rangers are currently in second in the AL West in playoff position and we lead them by two games but BR has them at 19.4%, while the third place in the AL West Astros are at 93.3%.

The AL team that has the 9th best odds of making the playoffs per Baseball Reference is the Blue Jays at 32.5%, and we have 3 games on them.

Maybe they are just looking at run differential? Orioles are +16 - while the Tigers are -2 and the Rangers are -14.

It's still so early, even in this mini season. The O's are 11-7, things can swing pretty quickly.

[Filmstudy]

9 hours ago, DrungoHazewood said:

A little real life data to see how it compares to the probabilities:

There have been 16 MLB teams who finished with a winning percentage between .320 and .329.  Over their first 17 games they ranged from 3-14 to 8-9.  Overall they were 88-167, good for a .345 winning percentage.  So nobody who finished with a winning percentage in the .320s ever started as well as 10-7.

There have been 20 teams that finished between .330 and .339.  They ranged from 10-7 to 0-17 (that being the 1988 Orioles).  The 1908 NY Highlanders, soon to be Yanks, went 10-7 on their way to 51-103.  Two other teams were over .500 in their first 17 games on their way to 100+ losses, I believe those were the '27 Red Sox and '54 A's.  All of these .330 teams totaled 110 wins and 230 losses, good for a .324 percentage.

So it is unlikely but not unprecedented for a .330-ish team to start with a winning record over their first 17 games.

I appreciate you researching, but you should certainly expand the range in both directions, not try to find the first wining percentage at which there was a 5% occurrence of a 10-7 start.  That's statistical cherry picking. 

Also, do you think it's compelling that 1 of 46 teams that finished with a winning percentage between .320 and .339 began the season 10-7?  That observed probability is actually 2.2%, lower than the 2.3% expected from random binomial draws of a .325 team.  If we assumed .330 was the average win percentage of the group, we'd expect 2.5% to start with a record of 10-7 or better with games decided by random draws.

DH, you obviously have some understanding of the math involved here.  What level of statistical significance are you looking for?  And how would you propose modeling in what we know from observing the 2020 Orioles to beginning-of-season expectations? 

[Frobby]

I’m loving the math talk, and lamenting that my once-excellent knowledge of probability calculations has gone to rust.    

I’m coming to the conclusion that it’s better not to pre-judge the “true talent” of the team, and just enjoy what’s happening and root like hell for it to continue.   It’s mid-August and we’re in the race — so what if we’ve played 18 games and not 120?    Just keep posting the W’s!
 

[now]

1 hour ago, Frobby said:

I’m loving the math talk, and lamenting that my once-excellent knowledge of probability calculations has gone to rust.    

I’m coming to the conclusion that it’s better not to pre-judge the “true talent” of the team, and just enjoy what’s happening and root like hell for it to continue.   It’s mid-August and we’re in the race — so what if we’ve played 18 games and not 120?    Just keep posting the W’s!
 

Yep, while the science can get ever more debatable, there's always good ole religion!

[wildcard]

The math goes out the window if Elias trades Cobb and Givens by August 31st.   Which I think is possible if he projects this team to be a .500 team that goes nowhere in the playoffs.  That is why the O's have to play .570 baseball for the three week period prior to August 31st.  To convince Elias that they are better than a .500 team.  JMO