Almunia vs Schwarzer: A Statistical Comparison

With all the transfer rumors surrounding Arsenal and Mark Schwarzer, it's time to take a look at Arsenal's current number one and his potential replacement. All data comes from a variety of public sources, but it is ultimately tracked by Opta.

Almunia vs Schwarzer
2009 Season
	Almunia		Schwarzer
Minutes	2610		3330
Shots/ 90 min	3.3		4.6
Save % In Box	55.6%		64.4%
Save % Out of Box	90.9%		92.9%
Catch %	66.2%		70.6%
Punch %	24.6%		22.2%
Drop %	9.2%		7.1%

Using the 2009 season to gauge the difference between Schwarzer and Almunia, we see that Schwarzer bettered Almunia in every above statistic. Schwarzer faced about a shot and a half more per game than Almunia while also having a better save percentage. While these numbers don't tell us everything we need to know, such as shot quality and type of scoring chances, we can make general assumptions based on the above numbers.

Almunia saved just 55.6% of the shots he faced in the box compared to 64.4% by Schwarzer. Almunia’s save percentage would rank 17th out of 21 keepers who played at least 1000 minutes last year while Schwarzer ranked ninth. Marcus Hahnemann of Wolves was tops in the league with an in box save percentage of 72.2% and David James was worst at 51.2%.

The final three stats, Catch %, Punch % and Drop % all deal with crosses. Almunia dropped 9.2% of the crosses he faced, the second worst for keepers with at least 1000 minutes in the premier league last year. While drops are nothing new for Almunia, he dropped just 4.8% of the crosses in 2008 compared to 3.6% by Schwarzer in the same campaign. Over the last two seasons, the average drop percentage in the Premier League is 4.6%.

The Kids Are Alright: Evaluating Goal Scoring By Age

Let's take a look at scoring in the English Premier League over the past two seasons. The goal is to define a peak scoring age in the league while also showing just how different the top three clubs are in their age philosophy.

The below chart shows scoring in the EPL since 2008 based on player age. The blue line is the actual goal totals while the purple line shows what we should expect when we regress the data. The reason I have done this is to show that the lack of goals from 26 year old players is simply a statistical oddity and not some unique trend.

As we can see, goal scoring peaks between 25 and 27 years old. Virtually no goals are scored once players hit 33 in the same way that very little goals are scored by players 20 and under. However, let's take a look at Arsenal over the past two seasons to see just how different they are from the rest of the league.

Here I have made a chart that shows Arsenal goals compared to the rest of the EPL. I used percentage of team goals here since the raw numbers don't really work for graphical purposes.

The above chart shows the dramatic differences in goal scoring by age between Arsenal and the rest of the league. Arsenal's goal scoring prowess starts at a much younger age than the other teams. Arsenal's goal scoring starts around age 19 before peaking around 22 while having virtually no contributions after age 29. For most teams in the Prem, the goalscoring gradually climbs from age 18 to 25, but Arsenal's goal scoring record shoots up drastically between ages 19 and 22.

Another interesting chart to examine is the differences in goal scoring age between Arsenal, Manchester United and Chelsea. We know that their philosophies differ often, but the below chart shows just how different the three clubs are when it comes to player age and experience over the last two seasons.

There are three unique philosophies at play here. First we see Arsenal, which has the smallest peak between the three clubs while also starting the earliest. Quite simply, Arsenal spreads the goals out over a greater number of players while having a much younger core. This should not come as a surprise to Arsenal supporters.

The next team to examine is Manchester United. United has two peaks, one at age 23 and 24 (Wayne Rooney's years) and what between 27-28 (Berbatov). Outside of these two peaks, there are virtually no other quality goal scorers. Manchester united has gotten no contributions in the goal column over the past two years from players aged 16-22 and 29-33.

Chelsea, the most recent Premier League champions, have one large peak that starts at age 29 and finishes at 31. This peak consists of Lampard, Drogba and Anelka. These three players have combined for over 70% of Chelsea's goals in the past two seasons.

Looking at these three teams gives us three unique views on player age and their contributions to the score sheet. Arsenal starts the youngest while having the most spread out talent, Manchester United is reliant on two main players in their mid-20's and Chelsea receives most of their contributions from players who are going into their thirties.

The Big Four

Unquestionably, the four most important members of the Arsenal squad last year (and going forward) were Cesc Fabregas, Robin van Persie, Alex Song and Thomas Vermaelen. The Big 4, as they should be known, are an essential part of the Arsenal squad and their health could be a deciding factor in the 2010 season.

The below chart shows Arsenal's record in 2009 when The Big 4 played.

The Big 4	W	D	L	Gm	Pts	Pts/Gm	Win %
All 4 Played	7	1	1	9	22	2.4	77.8%
Missing 1	7	1	5	13	22	1.7	53.8%
Missing 2	7	3	1	11	24	2.2	63.6%
Missing 3	2	1	2	5	7	1.4	40.0%

Arsenal played just 9 out of 38 games with The Big 4 all involved, but they averaged the most points per game when they were all involved. Arsenal averaged 2.4 points per game when The Big 4 played, but just 1.8 points per game when at least one member was hurt. Robin van Persie's long term injury certainly makes the number of games missed more than it should be, but the fact that Arsenal still played five matches with only one of their four most important players involved is a cringe worthy thought. All five of those games where three members of The Big 4 were missing were played without Song and Fabregas, the two essential members of the midfield.

Arsenal Fixture Analysis: 2006-2009

Looking back on past season's successes and failures can help us to better understand the future. Today we look at five trends over the past four years in regards to Arsenal's fixtures.

1. Double Trouble: In 2009, eight of Arsenal's nine losses came in back to back matches. On four separate occasions, Arsenal lost back to back matches. Since 2006, Arsenal had only done that four other times before (3 in 2006, 1 in 2008). Two of the four occasions that contained back to back losses started with losses to Manchester United. Arsenal must be more aware of this hangover effect for 2010.

2. Slow Starts Sink Arsenal: While Arsene Wenger loves to talk about November being the toughest month for his squad, it seems that August and September might be the most important. Over the last four seasons, the only Arsenal squad to reach 80 points came in 2007 when the Gunners finished with 83. In 2007 Arsenal took 19 points in seven games, their best mark by far. While Arsenal only played six games in their first two months in the other seasons, the 2.7 point average is still far and away tops.

3. Oh, November: November truly is an awful month for Arsenal and the stats back up Arsene's belief. Arsenal teams over the past four years have won just 38% of their matches in November while averaging just 1.3 points per game. Most troubling for Arsenal is that their points per game seem to peak in January before falling off over the final four months of the season.

Month	W	D	L	Win %	Pts/Gm
Aug	6	2	3	54.5%	1.8
Sep	11	1	2	78.6%	2.4
Oct	9	5	0	64.3%	2.3
Nov	6	2	8	37.5%	1.3
Dec	15	8	2	60.0%	2.1
Jan	10	5	1	62.5%	2.2
Feb	6	5	1	50.0%	1.9
Mar	9	4	3	56.3%	1.9
Apr	9	5	4	50.0%	1.8
May	5	3	2	50.0%	1.8

4. Any Day Will Do: Arsenal don't show any significant tendencies for a specific day of the week, but over the past four years they have tended to play better on Saturdays and worse on Sundays.

Day	W	D	L	Pts/Gm
Sat	50	22	13	2.02
Sun	22	9	10	1.83
Mid-Week	14	9	3	1.96

5. A bad trend: Arsenal have struggled to win matches decided by one goal. Since 2007, their win percentage has gone downhill from 77% in 2007 to just 50% last year. 2007 coincidentally was their highest season point total.

One goal Games
Since 2006

Year	W	L	Win %	Pts/Gm	Matches
2006	9	6	60.0%	1.8	15
2007	10	3	76.9%	2.3	13
2008	6	3	66.7%	2.0	9
2009	5	5	50.0%	1.5	10

As Arsenal get going with the 2010 season, we will check back on these five observations to see where the team is headed. Most important for Arsenal will be to get off to a good start through the first two months of 2010, in which they play three at home and three on the road.

Plus-Minus Follow Up

Just following up on yesterday's article regarding plus-minus. Plus-Minus is a useful tool to see how a player played in relation to the rest of his team, but should not be considered an official ranking of players on the roster.

In fact, soccer plus-minus must be taken with a large grain of salt because of sample size issues. In other sports, basketball and hockey, there are much larger sample sizes to work with because of the number of points accumulated in those games. When working with small sample sizes like the below article, we must understand that they numbers give us a small insight into player performance, but does not fully encompass a players impact on the team. Certain players are helped by situations out of their control while others suffer due to team circumstances that they play no part in. Let's take a look at Eduardo, who had questionable results from the 2009 season:


Eduardo: Eduardo's +/- per 90 minutes of 1.77 was tops among regular players for Arsenal in 2009. Simply put, Arsenal outscored their opponents by 1.77 goals for every 90 minutes Eduardo was on the pitch. Most Arsenal supporters would say how can that be? We all know Eduardo struggled to regain the form after the leg break in February 2008 and only netted twice in 24 appearances for Arsenal in 2009. Well Eduardo's strong +/- is due to circumstances of late match substitutions more-so than his individual abilities. Eduardo started 13 matches and came in as a sub 11 times. In the 13 matches that Eduardo started in, he played 945 minutes and Arsenal had a +/- of +16. In the 11 matches which Eduardo was a sub, he played 176 minutes and Arsenal had a +/- of +6 in that time. Look at this chart:

Eduardo	+/-	Mins	+/- per 90
Started	16	945	1.52
Sub	6	176	3.07

Eduardo had a +/- per 90 minutes of 1.52 while he was a starter, but an amazing 3.07 coming in as a late match sub. Eduardo's +/- has less to do with his own ability and more because of Arsenal's ability to score late goals. Without looking hard at other players, my guess would be strikers who come on as late subs will often see this happen. By virtue of playing a lot of late minutes, Eduardo was helped by Arsenal's ability to score at the end. Here is Eduardo's match log from the 2009 season:

Fixture	Result	+/-	Start/Sub	Min
1	W	0	Sub	73
2	W	3	Start	71
3	L	0	Sub	79
4	L	(1)	Sub	77
5	W	3	Start	74
6	W
7	W
8	W
9	D	0	Sub	88
10	W	3	Sub	37
11	W	4	Start	71
12	L	(1)	Start	71
13	L	(2)	Start	57
14	W
15	W
16	D	0	Sub	64
17	W	2	Start	75
18	W	0	Start	64
19	W	3	Start	72
20	D	0	Start	75
21	W	2	Start	84
22	W	2	Start	90
23	D	0	Start	62
24	L
25	L
26	W
27	W
28	W	2	Sub	84
29	W	1	Sub	74
30	W	1	Sub	76
31	W	0	Sub	84
32	D
33	W	0	Start	79
34	L
35	L
36	D
37	L	0	Sub	78
38	W

Introducing Arsenal Plus-Minus

This article is to act as a primer for the type of statistical analysis that is to come on this site and we will continue to look at Arsenal's plus-minus throughout the 2010 season.

Plus-Minus, or quite simply goal differential, is a statistical tool used to evaluate players. Plus-Minus is a common stat found in hockey, but rarely (if ever) used in soccer. The goal of the statistic is to determine how well a player played in the context of their team. Players with a positive plus-minus (+/-) saw their teams score more goals than they conceded while in the game. One of the most important concepts to grasp with plus-minus is just how important it is to adjust the numbers for team performance, because bad players on good teams will still look good while good players on bad teams will be brought down because of their teammates.

Stats Glossary:

For this article, the most important statistics are the final three: GD with, GD w/o and Net GD

Total: Arsenal goals minus opponent goals while player was on the pitch.
Minutes: Total minutes played on the season
+/- per 90: Total plus-minus per 90 minutes. Calculation is (P-M / minutes)*90
Adj +/-: Player +/- minus team's +/-
GD With: Arsenal's goal differential with player on the pitch
GD W/O: Arsenal's goal differential with player not on the pitch
Net: GD With minus GD W/O

Lets take a look at a few interesting players from 2009:

Thomas Vermaelen: Vermaelen led all Arsenal players in 2009 with a raw plus-minus of 40, meaning Arsenal outscored its opponents by 40 goals while Vermaelen was on the pitch. Vermaelen started 33 of Arsenal's 38 matches in 2009 and played in 82% of the minutes on the season, tops on the team. Vermaelen's +1.29 per 90 minutes was third best among players who played at least 1,000 minutes, but where we really see his effectiveness is in the Net Goal Differential. Arsenal outscored their opponents by 1.29 goals per 90 minutes while TV5 on the pitch, but that number dropped to just .29 when Vermaelen was out. That net difference of 1.00 was second highest on the team among regulars, outlining just how important Vermaelen was to the Gunners in 2009. There is no doubt how important Vermaelen will be in the future, highlighted by the fact that Arsenal took just five out of a possible 15 points while Vermaelen was sidelined towards the end of the 2009 campaign.

William Gallas: Vermaelen's center-back partner throughout much of 2009, Gallas started the first 26 matches before going down with injury for the rest of the season. While Gallas will be moving on to a new club in 2010, it's important to look back on his final season with the Gunners. Arsenal outscored its opponents by 1.19 goals/90 during the first 26 games while Gallas was active and while that pace dropped over the final 12, they were still .92 goals per game better than the opponent. This net differential for Gallas was just .28 goals, well below Vermaelen's 1.00 net differential. So while Gallas was an important figure for the Arsenal back line, his loss shouldn't have a major impact for next year.

Bacary Sagna: Although Sagna played in 35 matches, tops on the team, he also struggled at times in 2009 and was pushed for a starting spot by Emmanuel Eboue. Sagna had a plus-minus of just .95 per 90 minutes, well below the Arsenal team average of 1.11. Arsenal were .83 goals better when Sagna was on the bench and .29 goals better when Eboue was in the game. The RB position will be an interesting one to watch in 2010.

Denilson: Every Arsenal supporter seems to have a different opinion on the young midfielder, but this much is clear: Arsenal were a much better squad in 2009 when Denilson was on the bench. With Denilson in the game, Arsenal had a goal differential average of .63, compared to 1.51 when he was off the pitch. The -.88 difference was the second worse for any player who played in at least 1500 minutes (Nasri) and Denilson took part in five of Arsenal's nine losses in just 20 total matches played.