Just for fun, I added up all my teams' hitting, pitching and defense rankings. It's pretty clear what I valued most. Remember that a STDEV score of 0.0 means league average.

My 29 teams total standard deviation scores
Hitting: +1.7
Pitching: +25.0
Defense: +14.5

Average per team
Hitting: +0.1
Pitching: +0.9
Defense: +0.5

Clearly, all those late round unbalanced great-pitching-terrible-hitting teams skewed this total a bit. But when push came to shove, and the choice was between two fairly equal teams, I went pitching over hitting.

Posted by schwarze on 1/19/2023 8:19:00 AM (view original):

Posted by barracuda3 on 1/19/2023 7:09:00 AM (view original):

When juice announced the format for round 2 it was immediately apparent that research would be a herculean task. Creating 24x16=360 teams was just not feasible. Or was it?

Trying to figure out how you came to only 360 teams? I counted 700 possible teams. Did you build all 700?

You're correct. I actually drafted all 700. So it was like two workweeks rather than one.

Wow - you are a better man than I.

I built 243 teams, and 31 of these didn't get drafted, which means roughly 150 teams that I didn't build got drafted, including a handful of teams that were picked in the top 5 in their respective leagues. Can't wait to see which teams I ignored that will end up winning 90+ games.

Posted by schwarze on 1/19/2023 8:37:00 AM (view original):

Posted by barracuda3 on 1/19/2023 7:09:00 AM (view original):

Pick #15: League 2: 1901-1902 Philadelphia Athletics/Baltimore Orioles (A.L. West)
.334/.401/.499 (1), 2.62 (28)

Here’s where it starts to get interesting. The obvious draw here is the $23.7M 1901 Lajoie. .417/.461/.649, A+ at 2B and SS goes a long way. However, even if you remove him from the roster this team’s OPS# is still .872, which would rank second in this league. So yeah, my lineup is far and away the league’s best when everyone plays. But that won’t be all the time, as the ’02 Lajoie (.370/.418/.552) only has 459 PA and ’01 McGraw (.506 OBP) only has 376. And my pitching staff ranks only 16^th among teams that were drafted. On the plus side, my defense is pretty good. All of this to say, I’m not completely sold on this team. On paper, in theory, they’re pretty good. But I don’t know how it will play out. This was likely a reach. I wanted them so I could avoid having to be involved in the League 2 draft, but that was a high price to pay if I don’t make the postseason.

I think deadball teams are especially difficult to compare using overall stats. Clearly, I have the best offense and worst pitching in my division. I’m showing my primary contender as chisock’s ’07-’08 Pirates/Senators, who I have at .804/2.37 compared to my .900/2.62. But what does that mean? Is a point of OPS as valuable as a point of ERC? Probably not. But what’s the ratio? My gut feeling is that a point of ERC is worth twice as much as a point of OPS, which would suggest that all else being equal my team should still be better. But that’s a complete guess. If it’s worth 4 times as much then my team is worse. Anyway, I won’t be shocked if this team fails to make the playoffs.

I just thought of a way to attempt to quantify the value of a point of OPS vs. ERC. I wish I had thought of this before the draft. Now I’m afraid to do it because it probably turns out that most of my teams suck.

This is why I use standard deviations. By calculating the league average and standard deviation for ERC# and OPS#, you can see exactly how an unbalanced team fares. Also, each league is different. In a tightly packed league, the standard deviation is smaller, so an outlier offensive or pitching unit will be more valuable.

One of the things I noticed is that most of *my* high-ranking teams that dropped to me in the second half of the draft were unbalanced teams. One extreme example was the 1975-76 Dodgers/Padres (I took at pick #20). Based on the rosters I built, this team was a whopping +2.4 standard deviations better than league average in pitching (ERC#) but a horrific -1.9 standard deviations in offense (OPS#). Just summing up those two standard deviations rates this team as above average, and then after factoring in that they are +1.9 standard deviations better in fielding, this team jumped up pretty high in my rankings.

That's interesting. I never thought about using standard deviations, but they are yet another thing that Excel is good at calculating.

Out of curiosity, do you factor speed and/or stolen bases into your calculations?

Posted by schwarze on 1/19/2023 9:31:00 AM (view original):
Wow - you are a better man than I.

I built 243 teams, and 31 of these didn't get drafted, which means roughly 150 teams that I didn't build got drafted, including a handful of teams that were picked in the top 5 in their respective leagues. Can't wait to see which teams I ignored that will end up winning 90+ games.

Some might argue that makes me a worse man than you, but I tend not to listen to those people.

Posted by barracuda3 on 1/19/2023 9:11:00 AM (view original):

Posted by schwarze on 1/19/2023 8:19:00 AM (view original):

Posted by barracuda3 on 1/19/2023 7:09:00 AM (view original):

When juice announced the format for round 2 it was immediately apparent that research would be a herculean task. Creating 24x16=360 teams was just not feasible. Or was it?

Trying to figure out how you came to only 360 teams? I counted 700 possible teams. Did you build all 700?

You're correct. I actually drafted all 700. So it was like two workweeks rather than one.

Just once, I'd like that Herculean task. Usually when it gets to my pick, it's as easy as "Which pile of garbage looks most appealing?"

Posted by barracuda3 on 1/19/2023 9:41:00 AM (view original):

Posted by schwarze on 1/19/2023 8:37:00 AM (view original):

Posted by barracuda3 on 1/19/2023 7:09:00 AM (view original):

Pick #15: League 2: 1901-1902 Philadelphia Athletics/Baltimore Orioles (A.L. West)
.334/.401/.499 (1), 2.62 (28)

Here’s where it starts to get interesting. The obvious draw here is the $23.7M 1901 Lajoie. .417/.461/.649, A+ at 2B and SS goes a long way. However, even if you remove him from the roster this team’s OPS# is still .872, which would rank second in this league. So yeah, my lineup is far and away the league’s best when everyone plays. But that won’t be all the time, as the ’02 Lajoie (.370/.418/.552) only has 459 PA and ’01 McGraw (.506 OBP) only has 376. And my pitching staff ranks only 16^th among teams that were drafted. On the plus side, my defense is pretty good. All of this to say, I’m not completely sold on this team. On paper, in theory, they’re pretty good. But I don’t know how it will play out. This was likely a reach. I wanted them so I could avoid having to be involved in the League 2 draft, but that was a high price to pay if I don’t make the postseason.

I think deadball teams are especially difficult to compare using overall stats. Clearly, I have the best offense and worst pitching in my division. I’m showing my primary contender as chisock’s ’07-’08 Pirates/Senators, who I have at .804/2.37 compared to my .900/2.62. But what does that mean? Is a point of OPS as valuable as a point of ERC? Probably not. But what’s the ratio? My gut feeling is that a point of ERC is worth twice as much as a point of OPS, which would suggest that all else being equal my team should still be better. But that’s a complete guess. If it’s worth 4 times as much then my team is worse. Anyway, I won’t be shocked if this team fails to make the playoffs.

I just thought of a way to attempt to quantify the value of a point of OPS vs. ERC. I wish I had thought of this before the draft. Now I’m afraid to do it because it probably turns out that most of my teams suck.

This is why I use standard deviations. By calculating the league average and standard deviation for ERC# and OPS#, you can see exactly how an unbalanced team fares. Also, each league is different. In a tightly packed league, the standard deviation is smaller, so an outlier offensive or pitching unit will be more valuable.

One of the things I noticed is that most of *my* high-ranking teams that dropped to me in the second half of the draft were unbalanced teams. One extreme example was the 1975-76 Dodgers/Padres (I took at pick #20). Based on the rosters I built, this team was a whopping +2.4 standard deviations better than league average in pitching (ERC#) but a horrific -1.9 standard deviations in offense (OPS#). Just summing up those two standard deviations rates this team as above average, and then after factoring in that they are +1.9 standard deviations better in fielding, this team jumped up pretty high in my rankings.

That's interesting. I never thought about using standard deviations, but they are yet another thing that Excel is good at calculating.

Out of curiosity, do you factor speed and/or stolen bases into your calculations?

No, I do not factor in speed or stolen bases. Unless you're a Raines type base-stealer, I usually don't even bother with stealing bases. I will miss out of some teams b/c of this, like that mid 80's Raines x2 team that pedrocerrano selected.

And I know I need to improve how I factor in defense. Even in real life analytics, the proper evaluation of defense's affect on pitching has always been difficult to quantify. In theory, I should be applying the defensive grade to the pitching score (good defense improves my pitching score, bad defense makes it worse), but I've not figured out how to accurately do that, so I basically keep track of defensive score separately. So I add the pitching and hitting standard deviations together (i.e., equal weight), then add in a fraction of the defensive standard deviation to get my final roster score.

I really need to take some data from completed sim leagues and figure it out. This wouldn't be that hard to do, but the problem is that extreme ballparks complicates things. So, if a certain team's pitching is better or worse than expected (based on my formula), how much it due to defense vs playing in Petco or Wrigley? And of course, salary cap matters too. Even in $100M leagues, players will mostly underperform their real life stats... so how much weight should defense have on pitching in a $100M league a $160M league? A league with no cap?

Thinking about this some more, maybe I don't really want to solve how to accurately measure defense . The mystery is what keeps me coming back. Back in the early days of WIS, I remember when pitchers' salaries were 100% based on a player's real life ERA. Find pitchers who's ERA was way higher than their ERC, and you could crush the game. It was too easy back then, and I almost quit playing b/c of it. Luckily, admin recognized the flaw and recalibrated the salaries (this was well before the dynamic salaries).

Quote post by barracuda3 on 1/19/2023 9:44:00 AM:

League 10 - 1987-1988 Los Angeles Dodgers/Pittsburgh Pirates (A.L. Central)
.295/.371/.458 (27), 2.51 (8)

I have no recollection as to what I was thinking here. I don’t know why I though this was one of the five best teams in this league. Maybe I was right, but I’m currently clueless. I do remember agonizing between this team and the mid-80’s Expos with the two awesome Raines running seasons. Why I chose this one I haven’t the foggiest idea. And of course, this division filled fifth out of the six. But maybe that’s because people feared ribbentrop’s ’89-’90 A’s/Braves, who are likely the best team in the division. The very best explanation that I can come up with is that I attended Game 4 of the 1988 NLCS, which was the greatest baseball game I’ve ever seen in person. Oh well. I didn’t need 13 teams in the next round anyway.

I just set the lineup and settings for this team and perhaps my assessment was a bit harsh. I still don't think they're great, but maybe they have a chance to compete. The pitching staff is pretty good and well structured. And while the lineup isn't especially powerful, it has a lot of speed and efficient base stealers. I mean, the shortstop "platoon" (both RHH) is the Dave Anderson / Al Pedrique show, which is a program no one should want to watch (kind of like AfterMASH or The Ropers), but this team has sort of an endearing quality to it if you stare at it long enough. Spanky LaValliere shares catching duties with an elderly Rick Dempsey, who at least can reenact Babe Ruth's called shot on the wet tarp during rain delays. You've got the Bobby Bo cookie at 3B, Cangelosi as a part-time DH, The Lizard King who can indeed do everything (except catch) although nothing especially well, and when push comes to shove a hobbled Kirk Gibson can hit a clutch home run off of The Eck. In hindsight I'm kind of OK with these guys.

The 'Cude's write-ups should be required reading for all concerned. I think the only posts that made me laugh more were the JUGGALOSTEVE posts...lol.

Posted by barracuda3 on 1/19/2023 11:47:00 AM (view original):

Quote post by barracuda3 on 1/19/2023 9:44:00 AM:

League 10 - 1987-1988 Los Angeles Dodgers/Pittsburgh Pirates (A.L. Central)
.295/.371/.458 (27), 2.51 (8)

I have no recollection as to what I was thinking here. I don’t know why I though this was one of the five best teams in this league. Maybe I was right, but I’m currently clueless. I do remember agonizing between this team and the mid-80’s Expos with the two awesome Raines running seasons. Why I chose this one I haven’t the foggiest idea. And of course, this division filled fifth out of the six. But maybe that’s because people feared ribbentrop’s ’89-’90 A’s/Braves, who are likely the best team in the division. The very best explanation that I can come up with is that I attended Game 4 of the 1988 NLCS, which was the greatest baseball game I’ve ever seen in person. Oh well. I didn’t need 13 teams in the next round anyway.

I just set the lineup and settings for this team and perhaps my assessment was a bit harsh. I still don't think they're great, but maybe they have a chance to compete. The pitching staff is pretty good and well structured. And while the lineup isn't especially powerful, it has a lot of speed and efficient base stealers. I mean, the shortstop "platoon" (both RHH) is the Dave Anderson / Al Pedrique show, which is a program no one should want to watch (kind of like AfterMASH or The Ropers), but this team has sort of an endearing quality to it if you stare at it long enough. Spanky LaValliere shares catching duties with an elderly Rick Dempsey, who at least can reenact Babe Ruth's called shot on the wet tarp during rain delays. You've got the Bobby Bo cookie at 3B, Cangelosi as a part-time DH, The Lizard King who can indeed do everything (except catch) although nothing especially well, and when push comes to shove a hobbled Kirk Gibson can hit a clutch home run off of The Eck. In hindsight I'm kind of OK with these guys.

This is a solid team. I only have four team ranked ahead of them. It's rare that a team scores above average in all three categories, although it's slightly above average in both offense (+0.1 STD) and pitching (+0.3 STD). But they are a good defensive team (+0.8 STD). For somebody who claims they don't pay attention to defense, you sure do have a lot of good defensive teams.

Posted by schwarze on 1/19/2023 12:07:00 PM (view original):

Posted by barracuda3 on 1/19/2023 11:47:00 AM (view original):

Quote post by barracuda3 on 1/19/2023 9:44:00 AM:

League 10 - 1987-1988 Los Angeles Dodgers/Pittsburgh Pirates (A.L. Central)
.295/.371/.458 (27), 2.51 (8)

I have no recollection as to what I was thinking here. I don’t know why I though this was one of the five best teams in this league. Maybe I was right, but I’m currently clueless. I do remember agonizing between this team and the mid-80’s Expos with the two awesome Raines running seasons. Why I chose this one I haven’t the foggiest idea. And of course, this division filled fifth out of the six. But maybe that’s because people feared ribbentrop’s ’89-’90 A’s/Braves, who are likely the best team in the division. The very best explanation that I can come up with is that I attended Game 4 of the 1988 NLCS, which was the greatest baseball game I’ve ever seen in person. Oh well. I didn’t need 13 teams in the next round anyway.

I just set the lineup and settings for this team and perhaps my assessment was a bit harsh. I still don't think they're great, but maybe they have a chance to compete. The pitching staff is pretty good and well structured. And while the lineup isn't especially powerful, it has a lot of speed and efficient base stealers. I mean, the shortstop "platoon" (both RHH) is the Dave Anderson / Al Pedrique show, which is a program no one should want to watch (kind of like AfterMASH or The Ropers), but this team has sort of an endearing quality to it if you stare at it long enough. Spanky LaValliere shares catching duties with an elderly Rick Dempsey, who at least can reenact Babe Ruth's called shot on the wet tarp during rain delays. You've got the Bobby Bo cookie at 3B, Cangelosi as a part-time DH, The Lizard King who can indeed do everything (except catch) although nothing especially well, and when push comes to shove a hobbled Kirk Gibson can hit a clutch home run off of The Eck. In hindsight I'm kind of OK with these guys.

This is a solid team. I only have four team ranked ahead of them. It's rare that a team scores above average in all three categories, although it's slightly above average in both offense (+0.1 STD) and pitching (+0.3 STD). But they are a good defensive team (+0.8 STD). For somebody who claims they don't pay attention to defense, you sure do have a lot of good defensive teams.

I targeted this team in large part based on my experiences managing the 87 Pirates and 88 Dodgers to the WS in the Cooperstown League. You see a lot of weakness if you stare too long, but the reality was both these squads had deep pitching and found ways to surprise me consistently. I am often swayed by these odd emotional attachments, but too much substantive analysis can wreck the brain sometimes.

Fun fact: I was the one who wound up with those 89-90 Reds-Astros you mentioned way up high. At pick 16 even. I'm hoping your assessment of them was accurate!

PS: Getting to use Cangelosi in a DH league is a real treat. He's gonna set that table nicely.

Afterthought from readings 'cuda's always-entertaining breakdowns and schwarze's responses ...

My Excel skills suck. I'm basically using my Sheets as data repositories and not calculating anything. It's not that I don't know how to use these functions, but I've yet to think through how to use them effectively like this for these purposes. I'm obviously leaving some useful information on the table.

schwarze, how about your "10 Most Head-Scratching Draft Choices of the 2nd Round". I want to see how many spots I get in that Top 10, and we can see how accurate your choices pan out.

Oh man, I don't want to call anybody out just b/c I think they made a pick that I wouldn't have made. Were there selections made in the top 5 in a league that I show as below average teams? Yes. Am I going to name them? No.

However, there is one team I am really curious about. Before I started on the spreadsheets, I was just poking around different leagues, trying different builds. One of the teams that I looked at, and also was a team I was interested in, was the 1979-80 Royals/A's (picked 3rd by ybjsports in League 9). I really love that defense (+3.2 STD). When I got around to the spreadsheet calculations and plugged in the numbers, sadly the pitching rates -1.2 STD and the offense rates -0.4 STD, making this team a below average team, even with the defense bump. This team's success or failure will go a long way toward me understanding how much to weight defense.

Regarding your teams steveizzy, you had mostly lower round picks, so as it turns out, I only built 2 of your 6 teams. I have your "2009-2010: Florida Marlins - Detroit Tigers" team rated as below average, which isn't all that shocking as the 15th selection in league 13. I have your "1951-1952: Chicago White Sox - Boston Braves" (league 6, pick 21) as slightly below average.

Thanks, that makes sense as I am a slightly below average player.