Sports Reference Blog

Archive for the 'Advanced Stats' Category

WNBA Season Advanced Stats Now Available on Basketball-Reference

19th June 2018

As the 2018 WNBA season continues on through the summer, we have now added an Advanced Stats section to the season stats pages. We already included advanced stats on players' individual pages, so you already knew Brittney Griner had a 10.2 block percentage during her 2015 Defensive Player of the Year performance, but now you can see her performance with the context of the rest of the league. You could also take a look at the gap in win shares between Nneka Ogwumike's 2016 MVP season and the rest of her peers. Of course, we'll continue to update 2018 advanced stats along with the classics as this season unfolds.

Posted in Advanced Stats, Announcement, Basketball-Reference.com, Data | No Comments »

A New Chapter at Sports Reference

1st April 2018

I've been at this for 18 years now. Baseball-Reference.com launched in early April of 2000, and 18 years is a long time. We've gone from pre-formatted plain text tables to html tables, through various logos, and several redesigns involving mobile designs. Along the way, the only consideration has been what would work best for the user. That's why, after months of extensive research, I'm very pleased to announce:
Read the rest of this entry

Posted in Advanced Stats, Announcement, Ridiculousness, Tips and Tricks, Tout, Uncategorized | 25 Comments »

2018 WAR Update

15th March 2018

As you browse Baseball Reference, you might notice some subtle changes to WAR figures on the site. There are four main reasons for this:

  1. Park factors for recent seasons have been re-computed to be three-year rolling averages. For instance, 2016 Park Factors now encompass 2015-2017. This is something that needs to be done when seasons end.
  2. We've incorporated restated and expanded fielding statistics from Sports Info Solutions. SIS's Defensive Runs Saved forms the basis for our Defensive WAR calculations since 2003. From 2011 on they recalibrated data using their timer measures to measure ball hang time. There was also some recalculation based on changing shift methodology. Though we're now publishing their catcher framing stat (Strike Zone Runs Saved), we have not incorporated it into WAR at this time.
  3. Pitchers who received time as position players (whether PH or in the field) are now being treated as part-time pitchers and part-time position players. Previously we treated them as full-time pitchers. Some pitchers like Red Ruffing, Bob Lemon and Jim Kaat appeared in many games as a position player, pinch hitter or pinch runner. We used to credit these PAs as pitchers, which overvalued their offensive contributions. To handle this, we compute a percentage of time as a non-pitcher and make an adjustment.
  4. Also, we have incorporated a good deal more Retrosheet data which has affected the years we can compute more advanced fielding and baserunning measures. We're now able to roll these measures back to 1953. Another important change is that with Retrosheet gamelogs back to 1908, we can now use their IP data back to that year to get starter/relief IP splits. Some pitcher WAR changes for 1908-12 are due to WAR now being calculated using gamelog IP rather than the "official" total listed on the player's stat line. The biggest difference here was the appropriately named Bugs Raymond. The "official" record credits him with 324.1 IP that season, but the gamelogs come out to 304.1, which significantly impacted his WAR calculation (see below). For further reading on discrepancies between "official" records and more recently produced gamelogs, please read this excellent explainer by Retrosheet's Dave Smith.

For further details on WAR and its calculation, please see this WAR explainer.

Read the rest of this entry

Posted in Advanced Stats, Announcement, Baseball-Reference.com, Data, Statgeekery, Uncategorized, WAR | 8 Comments »

Pitch Framing Measures Added to Baseball Reference

8th March 2018

Our friends at Sports Info Solutions (formerly known as Baseball Info Solutions) have provided us with a pitch framing measure back to 2011, which we have added to Baseball Reference. Before I explain any further, if you're unfamiliar with the concept of pitch framing please read Mike Fast's 2011 article on the topic and Ben Lindbergh's 2013 follow up.

The stat that we have added is called Strike Zone Runs Saved. It represents the runs saved by catcher framing. In our tables, it's labeled RszC and it's available from 2011 to the present. While this statistic is a potential component of Defensive Runs Saved (and therefore WAR), please note that we have elected to not integrate this number into DRS (or WAR) at this time. We may elect to do so in the future, but for now we agree with Bill James's stance that waiting for further research is a good idea.

Read the rest of this entry

Posted in Advanced Stats, Announcement, Baseball-Reference.com, Data, Features, Statgeekery, WAR | 3 Comments »

Era-Adjusted Stats Added to Hockey Reference Player Season Finder

15th January 2018

At Hockey Reference, we calculate adjusted stats in order to more accurately compare players from different eras and different goal scoring environments. With Alex Ovechkin approaching his 600th career goal, there has been some chatter about Ovi being the greatest goal scorer in NHL history.

So we thought this made for a good time to introduce the adjusted scoring statistics into the Player Season Finder tool in order to test this hypothesis. As it turns out, Ovechkin is easily the NHL leader in "adjusted goals" through 13 seasons of play. If you click on the link, you'll see he has 691 adjusted goals (through 1/15/18), while Wayne Gretzky had 607 adjusted goals through 13 seasons.

Read the rest of this entry

Posted in Advanced Stats, Announcement, Data, Features, History, Hockey-Reference.com, Play Index, Statgeekery | Comments Off on Era-Adjusted Stats Added to Hockey Reference Player Season Finder

Provisional 2017 AV Live on Pro-Football-Reference

9th January 2018

With last week's announcement of the 2017 All-Pro teams, we're pleased to report that we've added 2017 Approximate Value (AV) numbers to the site for all NFL players. Note that these numbers are just provisional right now; the final numbers will be released after the Pro Bowl rosters are finalized. However, there's already some interesting preliminary information to dig through.

As of right now, Tom Brady and Todd Gurley are co-AV leaders at 19. For Brady, this represents the best season in NFL history by a player in his 40s. Gurley, meanwhile, is the youngest RB to hit that mark since Steven Jackson in 2006.

Not sure what AV is? To learn more about PFR's attempt to put a single number on each player-season since 1960 (for the purposes of comparing players across position and era), check out this link. For this year's full AV leaderboard, go here.

Posted in Advanced Stats, Announcement, Pro-Football-Reference.com | Comments Off on Provisional 2017 AV Live on Pro-Football-Reference

A Discussion of WAR Wherein I Ardently Attempt to Avoid any WAR-Related Puns

21st November 2017

This article assumes a lot of prior knowledge about the discussion of Wins Above Replacement, you can catch up here

First off, none of us are here without Bill James. We are all at our very best merely Chaucer or Joyce to his Shakespeare. All sabermetrics predating him flowed into his work and all sabermetrics after him carries echoes of his work.

To the discussion at hand.
Read the rest of this entry

Posted in Academics, Advanced Stats, Baseball-Reference.com, Stat Questions, Statgeekery, Trivia, WAR | 9 Comments »

Neutralized Baseball Stats Back and Better Than Ever

26th May 2017

When we launched our new site three months ago, one of the casualties was the neutralized stats tables we printed on player pages. Those tables were originally built using code that had grown increasingly unwieldy and were not something we'd be able to adequately maintain moving forward. That said, we're happy to announce that we have rebuilt these tables and added new features, as well.

Read the rest of this entry

Posted in Advanced Stats, Announcement, Baseball-Reference.com, Data, Features, History, Play Index, Redesign, Statgeekery | 3 Comments »

2017 NCAA Tournament Forecast

14th March 2017

Based on 5,000 simulations, here is our 2017 NCAA Tournament Forecast. We have Gonzaga as the likeliest champion, and 10-seed Wichita State as the 14th-likeliest champ. Don't forget to check this out before you fill out your bracket!

Posted in Advanced Stats, Announcement, CBB at Sports Reference, Features | Comments Off on 2017 NCAA Tournament Forecast

Russ & Harden’s Oscar-Worthy Performances

5th January 2017

We all know Oscar Robertson became the only player in NBA history to average a triple-double when he averaged 30.8 PPG, 12.5 RPG and 11.4 APG in 1961-62. Though he didn't have the benefit of first-class air travel, modern nutrition and training techniques, he did have the benefit of playing for a team that averaged an estimated 124.9 possessions per game, while playing over 44 minutes per game. Though I don't want to diminish the stamina required to perform at a high level at such a frenetic pace, the point stands that there were more opportunities for a player to accumulate counting stats in 1962 than there are in 2017, where the average game has about 96 possessions.

Still, even if we adjust statistics to a Per 100 Possessions basis, few have ever matched Oscar's production. Robertson averaged an estimated 26.7 Pts, 10.8 Reb & 9.9 Ast per 100 possessions in 1961-62. At Basketball-Reference, we have estimated per 100 possessions statistics for every player back to 1973-74 (when the advent of tracking of offensive rebounds, among other stats, made the estimates more reliable). From 1973-74 to 2015-16, a total of just four player seasons (by three players) matched The Big O's Per 100 Possession statistics. They were by do-it-all forwards Grant Hill & LeBron James (twice) and then Russell Westbrook, last season. But, now, the 2016-17 NBA season has TWO players doing that. Russell Westbrook, in his most ridiculous form yet, and James Harden, who has emerged as the perfect player for Mike D'Antoni's relentless attack. But it's not that these guys are matching Robertson's production. They're obliterating it.

Let's first consider Westbrook's numbers.

And yet, saying that Westbrook is like a combination of peak Jordan's scoring with peak Malone's rebounding and peak Kidd's passing doesn't even do justice to how breathtaking his level of activity on the floor is.

As for Harden, the numbers aren't as eye-popping as Westbrook's, but his scoring and rebounding are roughly equivalent to peak LeBron James, except with 50% more assists thrown in. So, I guess that's pretty decent.

Query Results Table
Per 100 Possessions
Player Season Age Tm Lg PTS TRB AST
James Harden 2016-17 27 HOU NBA 37.8 10.9 15.9
Russell Westbrook 2016-17 28 OKC NBA 44.1 15.0 14.9
Russell Westbrook 2015-16 27 OKC NBA 33.9 11.3 15.1
LeBron James 2012-13 28 MIA NBA 37.5 11.2 10.1
LeBron James 2008-09 24 CLE NBA 40.8 10.9 10.4
Grant Hill 1996-97 24 DET NBA 30.9 13.0 10.5
Provided by Basketball-Reference.com: View Original Table
Generated 1/5/2017.
Tonight, the former teammates will square off for the 13th time in their regular-season careers, with each player owning six wins to this point. Harden is currently the overwhelming favorite to the win the MVP Award, but it's still early. Anyways, with their matchup tonight, we just wanted to take the opportunity to gawk at their absurd 2016-17 seasons.
Top Candidates Table
Rk Player Tm W/L% eFG% FT% TRB AST STL BLK TOV PTS Prob%
1 James Harden HOU .750 .528 .847 8.2 11.9 1.4 0.3 5.7 28.4 57.6%
2 Kevin Durant GSW .861 .594 .868 8.6 4.7 1.2 1.6 2.3 25.8 16.1%
3 LeBron James CLE .765 .559 .683 7.8 8.5 1.4 0.6 3.9 25.8 8.8%
4 Russell Westbrook OKC .583 .462 .821 10.5 10.5 1.4 0.3 5.3 30.9 5.2%
5 Stephen Curry GSW .861 .575 .929 4.3 5.8 1.7 0.1 2.8 24.2 3.0%
6 Kawhi Leonard SAS .800 .530 .918 5.9 3.1 2.0 0.6 1.9 24.0 2.9%
7 Chris Paul LAC .632 .543 .877 5.3 9.5 2.2 0.2 2.3 17.7 1.9%
8 Kyle Lowry TOR .676 .585 .854 4.9 7.1 1.4 0.3 2.9 22.2 1.6%
9 Giannis Antetokounmpo MIL .529 .559 .779 9.1 5.8 1.9 2.0 2.9 23.9 1.5%
10 Kevin Love CLE .765 .534 .872 10.9 1.7 1.1 0.4 2.1 21.8 1.3%
Provided by Basketball-Reference.com: View Original Table
Generated 1/5/2017.

Posted in Advanced Stats, Announcement, Basketball-Reference.com, Data, History, Ridiculousness, Statgeekery | 1 Comment »