- This topic has 47 replies, 8 voices, and was last updated 2 years, 2 months ago by
.
-
Posts
-
This past offseason, the Cardinals traded “depth” in the outfield for relievers. The result has been the Cardinals have been caught shorthanded in the outfield, but have had a pretty impressive bullpen.
At least one commenter has stated they prefer position players that play 9 innings every day to relievers that play one inning every 2-3 days.
So, I thought it was worthwhile to see how bWAR is calculated and whether it is a measure of a player’s value.
As of this morning, the Cardinals have played 33 games into the season, just over one-fifth of the way through.
Players traded away:
– Tyler O’Neill: 1.0 bWAR in 184.1 innings in the field
– Richie Palacios: 1.1 bWAR (increased 0.1 since yesterday) in 191.1 innings in the fieldPlayers acquired:
– Andrew Kittredge: 0.7 bWAR in 14 innings pitched (increased 0.1 since yesterday)
– Nick Robertson: 0.1 bWAR in 4 innings pitchedPart 1 of 3
So, I thought it was worthwhile to see how bWAR is calculated and whether it is a measure of a player’s value.
Is relief pitcher WAR sophisticated enough to capture ‘inherited runners allowed to score’? For example, Kittredge allowed two inherited runners to score yesterday, without an error involved, but got a zero ERA for the outing. He was tagged with a blown save, according the the box score.
Let’s look at relievers first:
Andrew Kittredge: 14.0 IP; 1 Earned Run; 0.64 ERA – he pitched 2 innings last night, and his bWAR increase 0.1 to 0.7
Nick Robertson: 4.0 IP; 0 Earned Runs: 0.00 ERA
Both look phenomenal, correct?
But, as Bling pointed out with Robertson, he came in his last game and allowed 2 inherited runners to score. Similarly, last night, Kittredge allowed 2 inherited runners to score. On the season:
Kittredge: 4 inherited runners; 3 have scored (75%)
Robertson: 2 inherited runners; 2 have scored (100%)What if ERA for relievers was adjusted to account for inherited runners? We still have the same innings pitched, but instead of Kittredge’s bWAR increasing last night as his ERA decreased, his bWAR would have decreased as his “Real ERA” increased.
What would this “Real ERA” look like?
KIttredge: 14.0 IP; 1 Earned Run; 3 inherited Runs; 2.57 “Real ERA”
Robertson: 4.0 IP, 0 Earned Runs; 2 inherited Runs; 4.50 “Real ERA”Now, can we compute how many runs a “replacement” level reliever would have allowed? Not really, but we can compute how many an “average” reliever would have allowed.
As of today, the MLB average ERA for 2024 is 3.94. This implies:
14.0 IP (KIttredge) = 6.13 runs alloed (round to 6)
4.0 IP (Robertson) = 1.75 runs allowed (round to 2)So, on the season, using “Real ERA” instead of the generally accepted ERA which forgives relievers for allowing inherited runners to score:
Total Runs Saved:
Kittredge: 2 (1 Earned Run + 3 inherited runs)
Robertson: 0 (0 Earned Runs + 2 inherited runs)Part 2 of 3
Is relief pitcher WAR sophisticated enough to capture ‘inherited runners allowed to score’?
No, in my post above, I showed Kittredge was rewarded for a phenomenal outing, even though he allowed inherited runners to score.
His ERA went from 0.75 to 0.64. His bWAR increased from 0.6 to 0.7.
Finally, let’s take a look at the value of the position players traded away:
Tyler O’Neill: 184.1 innings; 13 RBI; 21 Runs Scored; 34 Total Runs; 1.0 bWAR
Richie Palacios: 191.1 innings; 6 RBI; 15 Runs Scored; 21 Total Runs; 1.1 bWAR (increased 0.1 since yesterday)Now, let’s look at the “Replacement” players:
Victor Scott II: 158.0 innings; 2 RBI; 6 Runs Scored; 8 Total Runs; -0.7 bWAR
Mike Siani: 162.2 innings; 1 RBI; 9 Runs Scored; 10 Total Runs; 0.1 bWAR
Alec Burleson: 60.0 innings; 10 RBI; 4 Runs Scored; 14 Total Runs; 0.0 bWARSo, the Cardinals traded 2.1 bWAR worth of outfielders for 0.8 bWAR of relievers (which is vastly overstated) and replaced the outfielders with -0.6 bWAR of replacement outfielders.
Net effect:
Outfield: -2.7 bWAR; -23 Total RunsPart 3 of 3
I will make a few more points:
As of this morning, the Cardinals have a -28 run differential.
– They cost themselves 23 runs by using “replacement” outfielders instead of actual outfielders
– They saved 2 runs with their bullpen acquisitionsThat run differential would be -7 instead of -28.
Second point, the Cardinals are 3 games under .500. They downgraded their outfield by -2.7 bWAR. So, the difference between 3 games under and a .500 team is the outfield.
If I were calculating WAR for relievers, Kittredge would probably be around 0.2 and Robertson would be at 0.0
No, in my post above, I showed Kittredge was rewarded for a phenomenal outing, even though he allowed inherited runners to score.
So WAR ignores both inherited runners allowed to score AND blown saves!
This is with the benefit of hindsight and also ignores surpluses and needs. In the off-season, it was generally seen that we had more OFs than roster spots for them while we also had a weak BP. While I was a fan of both O’Neill and Palacios, I understood the trades at the time. I also don’t think O’Neill would be playing nearly as well on this team with Marmol and the tensions there.
I doubt anyone here or anywhere predicted the Cardinals top 3 CF options in Edman, Carlson, and Nootbar would be in the IL at the same time. We knew about Edman, but not the extent and I doubt even the Cardinals FO knew the extent given their comments. Now I do think the Cardinals were way too aggressive with Scott as he should have been in AAA to start the season, but I lurked these boards and many were clamoring for Scott to make the team as well.
In general, comparing WAR in a trade isn’t a good measure as each team has different needs and those teams may be in different timelines of competitiveness. Sure RP don’t earn a lot of WAR, but without good RP your team can’t maintain leads consistently. It also becomes much more difficult to win in the playoffs where RP are even more vital.
Yes, it is with the benefit of hindsight.
But looking at instances where Cardinals have traded outfielders for relievers, it’s becoming apparent that trading position players for relievers is a bad way of building a team. The Cardinals front office should know this, if they do any competitive intelligence whatsoever. We, as fans, only follow one team. They have the duty to follow 29 other teams and build their models based on historical precedent.
And WAR is supposed to equate players at different positions. Clearly, WAR is failing with relievers.
It is also worth noting that O’Neill was this forums whipping boy the last 2 years. The majority wanted him gone while a few did want him to stick around given his tools. Fans are too eager to over react to both under performance and over performance. Sometimes patience is key.
Agreed. O’Niell plays 50% of the time, and was owed $5.5 million. But you trade him for prospects, not relievers.
The difference with Palacios was that he is making the league minimum, had 5 years of control and was traded for a reliever with one year of control.
Position players typically out-war relievers four to one (ballpark), so that baked in baseline does exist.
WAR is a great tool, but it isn’t the only tool in the box for creating a roster and sabermatricians will state that as well. WAR does a good job of comparing players and their impact. RP definitely do have less of an impact compared to full time players, although you still need them for a good roster.
WAR also tries to account for a single players contributions and not the team’s and thus tries to assume everything the player does in isolation. Some players in the past have looked much better just because the other players on the roster are great and they benefit from it. But if you want to truly evaluate a player, you need to know what he himself contributed.
That philosophy is why RP don’t get hurt for allowing inherited runners to score. In isolation, Kittredge did his job. Now for a team effort, it certainly hurt to allow those runs, but he was dealt a bad hand to begin with of 2 runners on base with no outs. On average, one of those runners is going to score in that scenario.
If you want to know what a player is contributing to the team’s overall chances of winning, I think you want to look at Win Probability Added instead.
Agreed. O’Niell plays 50% of the time, and was owed $5.5 million. But you trade him for prospects, not relievers.
It takes two to tango. O’Neill’s value had dropped significantly due to his poor play. Add on that we didn’t have leverage as other teams would have known about the “clubhouse” issues from Marmol thus making his value really low. I’m sure Mo tried to get prospects for O’Neill, but if no one was willing to give them up, he can’t force it. Now maybe the better move was to find a new manager and keep O’Neill, but we still do have a clogged OF in the off-season.
All, Fangraphs has a live box score daily, for following the game context, play-by-play.
Prepare to add it to your game routine. It’s that good.
Position players typically out-war relievers four to one (ballpark), so that baked in baseline does exist.
Not really. Kitredge has accumulated 0.7 WAR in 14 innings. At BEST, he has saved the Cardinals 5 runs. If you factor in inherited runners, he has saved the Cardinals 2 runs.
Meanwhile, Palacios has generated 21 runs. So, if it was 4X, and 5 runs saved equals 5 runs produced, Palacaios would be at 2.8 bWAR by now.
League, per season. 3 to 1 if you like. But not to get lost in the details, the unmentioned baseline is not 0 vs 0 in your comparison. To be fair…. I chose to mention it.
On average, one of those runners is going to score in that scenario.
Yes, they have a probability chart that states what the likelihood is of a runner scoring depending on which base he is on and how many outs there are.
So, rather than adding a full run for each runner to the ERA, you would add whatever that probability is.
It would be much more accurate than ignoring the inherited runs altogether.
League, per season. 3 to 1 if you like. But not to get lost in the details, the unmentioned baseline is not 0 vs 0 in your comparison. To be fair…. I chose to mention it.
It doesn’t really matter. There is 0% chance Kittredge has accumulated 0.7 WAR in 14 innings.
See, you made me go check myself 🙂
4.08 to 1 last season (BR).Kitt should have a good total, he’s been very good. That can also be .05 in a week. TBD
Not really. Kitredge has accumulated 0.7 WAR in 14 innings. At BEST, he has saved the Cardinals 5 runs. If you factor in inherited runners, he has saved the Cardinals 2 runs.
Meanwhile, Palacios has generated 21 runs. So, if it was 4X, and 5 runs saved equals 5 runs produced, Palacaios would be at 2.8 bWAR by now.
You are using the wrong tool and are basically trying to use a hammer to put a screw into the wall. Runs, RBIs, ERA, and inherited runners allowed are team statistics. WAR is trying to measure the individual players contributions. Even a mediocre player can look great on a bad team when measuring them by runs and RBIs. I’m not saying Palacios hasn’t played great as he definitely has been. But Kittredge himself has also pitched very well. Sure he allowed some inherited runners to score, but how many runners has he prevented over an average RP due to his ability to prevent runners to get on base?
So, rather than adding a full run for each runner to the ERA, you would add whatever that probability is.
Again, you are using the wrong stat. Exactly what you are asking for is available as a stat which is Win Probability Added. Palacios has a WPA (higher is better as it is a summation of % chance added to win probability) 0.61 and Kittredge has a WPA of 0.59. Palacios has been worth 2 more % chance of winning over the season than Kitty.
WAR is trying to measure the individual players contributions.
However, this isn’t the track and field, but rather a team sport. How many times last year did you see a Cardinal relief pitcher pitch effectively in garbage time, only to stink up the joint in a close contest?
Kittridge was asked to hold a lead, and he didn’t do it, and we eventually lost he game. Who cares what he does in a mop up role tomorrowIt doesn’t really matter. There is 0% chance Kittredge has accumulated 0.7 WAR in 14 innings
Kittredge has allowed 1 earned run over 14 innings for an ERA of 0.64. Assume the replacement level RP has a ERA of 4.5 on average (generous estimate), and they would have allowed 7 runs in that same amount of 14 innings. That is 6 runs of difference in high leverage situations where the game is on the line.
- You must be logged in to reply to this topic.