Just noticed something today...

[spikeslovechild]

When perusing BBRef I came across Marcus Semien. I knew he a had a ton of errors this season 31 to be exact what I didn't know is that his DWAR is actually positive.

How is that even possible? Does the scorekeeper in Oak hate him or something?

[LBJSeizedMyID]

If you watched the Jays game last week as an example you'd see that he's the worst shortstop in the MLB right now. That's was error #30 at the time

[joseph227]

Use fangraphs for war. They use UZR which is a better judge of defense.

[Skin Blues]

Errors are a very bad way to judge a defender. All else being equal, it's obviously better to not make errors, but you can easily make up for it with good defensive range and a strong arm. For a SS, a good defensive 1B can also save you from a lot of throwing errors.

For an example, lets look at Jimmy Rollins vs Alexei Ramirez for the past 5 seasons. Rollins has a better Fielding % (.983 vs .977) but lets look a bit deeper. On plays that are made about half of the time by MLB shortstops (40-60% conversion), Alexei Ramirez has converted an out 58.8% of the time. Jimmy Rollins, on the other hand, is at only 40.7%. Normalizing to 75 plays for arguments sake, Alexei makes 13 more outs than Jimmy. And that's just on one subset of plays. On plays converted between 10-40% of the time, Alexei makes 16 more outs than Jimmy. And so on. So while Jimmy might not be making errors, he's not converting batted balls into outs, either because he's slow getting to the ball, has a slow release, a weak arm, bad instincts - who knows why? Errors are only a small part of it.

Just for fun, here, lets look at Jose Reyes vs Brandon Crawford over the past 5 years. To simplify, we'll look at estimated total Defensive Runs Saved. It's a self explanatory stat; how many runs did you save your team (or cost your team) in a given time period? Jose Reyes had a .968 Fielding %, and Crawford .972 which is almost identical. It's a rounding error. However, over 5 years, Crawford has saved his team 105 more runs than Reyes.

As for Semien: it's a small sample size, first of all. And I think BREF uses only DRS to judge players, rather than hedging with UZR. By UZR, Semien is the worst SS in the league. By DRS he's above average. The biggest issue though, is sample size. Looking at the Inside Edge fielding numbers, he looks to be one of the worst in baseball across the board. I don't know what's up with DRS, but small samples can be funny.

Use fangraphs for war. They use UZR which is a better judge of defense.

It's better to use a combination of both, as the sample sizes are so small that neither on their own mean much in only one season of data.

[Schad]

Just for fun, here, lets look at Jose Reyes vs Brandon Crawford over the past 5 years. To simplify, we'll look at estimated total Defensive Runs Saved. It's a self explanatory stat; how many runs did you save your team (or cost your team) in a given time period? Jose Reyes had a .968 Fielding %, and Crawford .972 which is almost identical. It's a rounding error. However, over 5 years, Crawford has saved his team 105 more runs than Reyes.

Also known as the Derek Jeter Proof: far easier to avoid throwing errors if you make zero plays that require moving.

[Kurtz]

Use fangraphs for war. They use UZR which is a better judge of defense.

Fangraphs tells me Roberto Alomar was a bad defender as a Jay. That seem right to you?

[joseph227]

Use fangraphs for war. They use UZR which is a better judge of defense.

Fangraphs tells me Roberto Alomar was a bad defender as a Jay. That seem right to you?

UZR wasn't around before 2003.

[UN-Owen]

According to Baseball Reference: Alomar was -0.8 (dWAR) during his time with the Blue Jays