There are Celtic “fan” sites that lure you in with clickbait and sensationalist headlines, and there are those that simply regurgitate what others have produced, either the mainstream media or less forgivably, other Celtic fan sites.
I hope you realise I am never the latter and on a certain cheeky level, only occasionally the former.
The purpose of the above is to presage this small piece on Joe Hart. As it isn’t primarily about Hart (well, it is). I’m currently playing with a new toy to create radars and this was the first subject that came to mind.
With radars the art is in a) selecting the right data points and b) the right maximum and minimum boundaries.
I trust you appreciate that one of the (few, admittedly) benefits of this site is the ability to compare Celtic players over the years on a truly “apples to apples” basis. Most data sites provide radars in the context of the rest of the league (hugely distorted given the competitive inequalities of the SPFL Premiership) or limited to one competition.
I can show you how players compare against their peer group of other Celtic players and across all competitions.
Now the issue with goalkeepers is the population of custodians with more than 900 minutes since 2014/15 is 11! So, it’s a long game for this position.
Anyway, back to Hart.
Hart 2023-24
Here is how he is comparing to that small population of peers, this season.
Sorry the text is so small.
- GSAA = Goals Saved Above Average
- GSAxG = Goals Saved Above Expected Goals
- PEI% = Possession Efficiency Index
- Position Error = average starting distance away from “optimal” per shot faced
The rest should be explanatory.
What we maybe do not emphasise enough about Hart is how well he has adapted late in his career to be effective with his feet. Yes, he can look slow and uncomfortable on occasion, but he is a trier.
The bottom left corner of the radar is where the “distribution” metrics reside. By volume and result, he is the best goalkeeper Celtic has had since records began (!) in simply recycling the ball – a key component in modern play. His pack pass score is low meaning he isn’t particularly risky in his passing.
Otherwise, he is now at 36, less effective at keeping the bally out of the netty.
He is in negative territory as regards goals saved above (in this case below) average and in goals saved above (in this case below) expected goals. His simple save percentage is also in the 60s, not the preferred 80s.
Retirement is a wise course for him as he hasn’t yet fallen off the proverbial form cliff but is nudging in that direction.
Anyway, thanks for indulging and welcome feedback on the radar. More positions to follow.
Radars are a good way to show the complexity of comparing players. Although it’s a good way to highlight a player’s strengths and weaknesses, it’s sometimes a bit difficult to see what aspects of a player’s performance are “adequate”. Also, in a team game, players don’t need to be good at everything, so long as all the requirements are being fulfilled by someone eg you don’t need two aerially dominant CBs, but you probably do need one. I do like pairwise comparison radars. They can sometimes bring a bit of clarity to an analysis.
Hi Stephen
Yes i agree the comparison is a great use of them. My point here probably not well enough explained is that the max and min values are based solely on Celtic Keepers performance so that comparison is built in with a relevant peer group as opposed to eg Dundee’s keeper. Hope that makes sense. it is a first attempt too!
Really interesting stuff this and the stats appear to align with the performances – might just add at this point I have always been a real fan of Hart at Celtic.
We are in a tricky situation at Celtic – young GK’s are prone to concentration errors, and introducing a new one to the first team would be with immense risk particularly in the big games. Additionally, would they develop very much given 90% of their game time would be watching us in possession? Therefore loans for our younger keepers should surely be a must…
On the other hand, signing a peak GK for 5+ million…is it worth the money given how much we dominate domestically, would he only be for European games? And if so, would the money be worth it in this case? And would a peak international keeper be interesting in joining us playing most of his games as a spectator?
I suppose what I’m saying is that I prefer having a veteran in goal for us – experience, happy to marshall the nets in games we dominate, but can also step up for the odd big performance in Europe. I don’t think investing huge sums of money in a GK (in our particular case) will ever be a priority over forwards etc.
Really enjoy your site. Keep up the great work
Thanks Mark – very kind.
I think i’m settled on needing a keeper that will make improbable saves in important games ie Europe as being the most important differentiator. Forster was perfect in that regard and i would forgive the relative lack of ability with ball at foot.
Celtic need a keeper who is not completely hopeless with ball at feet. No more. We need a good shot stopper, ideally around 28/29 years old.
Had we signed and played But land, I think we would only have gained 1 extra point in the CL this year (draw with Feyenoord away, though this presupposes not being 1 nil down would have affected the red card situations). And we maybe would have lost fewer goals in Madrid, but still been soundly beaten.
I don’t think Hart has in and of himself cost us many points, but we need to change tact to a top shot stopper (or the best we can afford) and cross claimer above all else, as these are the areas that will make a material difference in European games.
Hart himself was historically awful with ball at feet and has really improved this since joining (honestly!) so that can be trained.
But agility, reflexes and decision making need to arrive with a signing when you’re a club like celtic. Otherwise the new guy won’t get a chance.
Side note, I would not welcome another return of Forster. He’s now at the steep downhill decline stage of a keeper’s career and heading to the cliff edge. That’s just age and completely natural. I’m also unconvinced by Kelleher. He looks like a keeper that will be good in 3 or 4 years as he matures. We don’t really want to be the club he spends those 4 years with. Maybe the 4 years after that…
Interesting article.
11 keepers playing over 10 games in 10 years seems a lot but perhaps the days of keeping for years on end without a break are more a thing of the past, especially in lesser and Cup games where some teams now elect to rest their top stopper.
Apart from the No. 1 No. 1s, Gordon, Foster and Hart, I assume Bain, Barkas and Zaluska would make the cut but not sure if Beni would jet into the top 11.
Can’t think of the other 4/5 but perhaps some youngsters or loanees who got emergency cover or lesser game time.
Looking at the radar then Hart underperforms, slighly, on shot saving but I wonder if there might be an “unfair” Celtic element here ie by and large our opponents tend to sit back and will only attack us when they have a significant chance of scoring.
Compare a dominant team like ourselves who will have far more attempts across the spectrum from poor to good chances.
It would be good to see a similar radar for the media darling, Butland. I don’t watch their games other than against us but my impression, from reports, seems to be that he’s under a lot more sustained pressure than Hart (who tends to only face the difficult to defend breakaways), which makes it easier to perform better if you’re reasonably competent.
Is he still essentially still at the level of the 3rd or 4th Palace choice behind ordinary keepers or has he, at this late stage, morphed into the great keeper that its a disgrace that Southgate hasn’t made England’s No. 1 and for whom, Rangers will graciously accept opening bids of £15m.
Are you aware of any Rangers data analysis sites that have analysed his performance using similar metrics?
Hi yes it wasn;t 11 individual keepers – sorry should have been clearer. It was 11 x instances of a keeper playing more than 900 minutes. It is Gordon x 5, Bain x 2, Forster x 1, Barkas x 1 then Hart x 3 – ie this season would be the 12th.
The comparative radar between Butland and Hart does not make good reading for Celtic fans. You can see it in this article – https://www.celticway.co.uk/opinion/24181005.celtic-rangers-three-key-title-race-factors-season/ – only cost you £1 currently per month!
TRFC actually give up far fewer chances than Celtic – the shot map against comparison would make you weep. it’s the biggest threat to the title.
On Statsbomb I’d classify Butland as like 1st season Hart for us. he is competent but probably 4th in league for key metrics like Goals Saved Above Average. He’s good but not as good as the frothing panderers would have it.
Gotcha, 900 mins in a season, so although guys like DeVries and Zaluska would have those mins over their careers, obviously not in a single season.
If Butland is facing less chances but doing better than Hart then that is worrying, even given our shared cynicism for our impartial media.
Who are the 3 that are ahead of Butland (Roos at Aberdeen gets a lot of good press but I don’t know enough whether its justified)?
I don’t know anything about alleged target Etienne Vaessen but my alarm bells ring when we’re after yet another free transfer, as for every Hart and Gordon, who were one-offs for different reasons, there are plenty of Bain, Siegrist, Zaluska, DeVries etc.
I hope I’m wrong but its almost as if the powers that be consider it a low priority position given how much we are dominant domestically but its key if we’re to advance in Europe.
Roos, Mitov, Dennis and unbelievably Marshall all higher than Butland for Goals Saved Above Average on Statsbomb. All have saved 2-4% more saves than expected indicating the tiny margins at play here. Dennis has the lowest Positioning Error average at 1.29. Roos is higher than Hart at 1.73 indicating he may be a bit “lucky”.
Goalkeeper should be the easiest position to recruit. Virtually all teams only have 1 main one meaning there are 2-3 at each club who want to be a number one.
I think a true reflection would have to include McGregor and Butland for our nearest challenger because they have challenged in the past few Seasons no doubt despite our apparent dominance trophy wise.
In particular McGregor who saved them countless times in Domestic and European competitions which truth be told they have done better than us over the past 5 years or so despite the increased level of competition where a keeper has to perform at a higher level as does the defence in front of him.
Also the defence this Season has changed on no less than 20 separate occasions out of the 30 league games Hart has played in so this is also a factor the stats don’t take into account.
It could be argued either way whether that has more or less significant effect on the keepers performance levels.
Talking of Radars though Celtics wealth unlike theirs has definitely been on the radar.
Given they haven’t posted a profit in 12 years of existence is it any wonder it was off the Radar ?