One of the top NHL unrestricted free agents this summer is the Washington Capitals Alexander Semin. Semin has seen his goal production drop from 40 goals in 2009-10 to 28 in 2010-11 to post-lockout low of 21 this past season and as a result peoples general view of Semin’s value has dropped significantly. The question is, what was the reason for his drop off in offensive stats. Is it Semin alone, or is there some other underlying reason.
Let’s take a closer look at Semin’s point totals over the past 5 seasons.
When you strip out Semin’s even strength performance you begin to realize that his point total drop off is not near as significant. The past 4 seasons he has had 47, 55, 45 and 43 even strength points. Now his time on ice between points increased dramatically this season but a significant part of that is likely due to his line mates. Three seasons ago Semin’s most frequent line mates were Brooks Laich, Nicklas Backstrom, Alexander Ovechkin and Tomas Fleishmann ordered by time on ice together. Two seasons ago they were Laich, Backstrom, Ovechkin and Johansson. This past season they were Johansson, Perreault, Chimera and Ovechkin. No offense to Johansson, Perreault and Chimera, but they have combined for just one 40 point season in their careers, and that was Johansson this past season playing with Semin. That certainly has a little to do with Semin’s drop off.
Another way to look at Semin is to take a look at how his team mates perform when they are on the ice with Semin and when they are on the ice without Semin. We can do this by looking at some nice bubble charts.
The above chart has GF20 without Semin across the horizontal axis and GF20 with Semin across the vertical axis. For those new readers, GF20 is goals for (i.e. scored by team) per 20 minutes of ice time. The color of the circle identifies the year and the size of the circle indicates relative ice time played with Semin. The larger the circle, the more minutes they played with Semin, the smaller the circle the fewer. Each forward who played at least 150 minutes with Semin are shown above.
In this chart circles in the upper left indicate that Semin had the greatest impact on his team mates offensive performance as upper-left circles indicate they performed relatively poorly without Semin and relative well with Semin. Anyone above the 1:1 diagonal line (not shown) means that they had a better GF20 with Semin than without. As you can see, over the past 3 seasons there is significant evidence that Semin has made his line mates better. That changed slightly this past season though. While Chimera and Perreault had better GF20’s with Semin, Johansson and Ovechkin did not.
Now lets take a look at the same chart but for GA20 (goals against per 20 minutes of ice time.
In this table bubbles in the lower right or below the 1:1 line are good as this indicates the player had a lower GA20 with Semin than without. Except for Ovechkin in 2010-11 the majority of the bubbles are pretty close to the 1:1 line or slightly below. This would seem to indicate that Semin is not a defensive liability which is relatively rate for quality offensive players. Frequently producing big offensive numbers comes at a cost of defensive performance but this does not seem to be true for Semin.
The final bubble chart I will look at is goals for percentage (GF%) which is simply goals for divided by goals for plus goals against.
GF% is like GF20, the higher the number the better, so like the GF20 bubble chart, bubbles in the upper left above the 1:1 line are better, especially if they are above 50% (i.e. more goals for than against). Except for Ovechkin and Johansson this past season and Morrison in 2009-10, all players had a better GF% with Semin than without. This clearly points to Semin having a significant positive impact on his teams performance.
Maybe the most impressive thing I can point out about Semin is his overall 2-way performance relative to the rest of the league. Of 125 players with 2500 5v5 zone start adjusted minutes of ice time over the past 3 season, Semin ranks 5th in GF20 (trailing only D. Sedin, H. Sedin, Toews, and Stamkos) and he ranks 13th in GA20. It truly is a rare combination (for example, the Sedin’s rank 28 and 38 in GA20, Toews 60th and Stamkos 105th).
All that said, it does appear that Semin had a slight drop off in 5v5 offensive performance this past season but without further evidence it would probably be fair to presume that that was a somewhat minor drop off from an otherwise exceptional 4 years and certainly wouldn’t be enough to scare me away from making a significant offer to him as an unrestricted free agent. He’d be a worthy addition to any NHL team.