Mayo GAA Discussion

Genuinely surprised you don't have more acknowledgement for this post. Very funny. Kudos

Running across a defender no bother when the player is in possession. Handy way to win a free. Doing it a million miles away from the action is a totally different situation trying to get a lad black carded. In the modern game players run side by side all game long tracking runners. Lads were throwing themselves into each like that no one would be left on the pitch.

Anyway my comment was in jest more in defence of AOS than anything else. Galway is one of the more frequent teams he battered playing full forward and I still get the sweats when he lines up there.

I dont think anybody said he isnt! He will have a league title in a few weeks as well im sure and good luck to him!

I see AhRef is running with The McStayWay as a follow on from HoranBall... Tongues firmly in cheek.

Yeah, that's a fun one to get your head around at a glance.

With Kerry favourites to beat Roscommon/Galway, on paper at least, they'll have multiple scenarios (for now) where they'd pip the other two on equal points. For the scenarios where Roscommon or Galway progress they'd need to have more points than Kerry so ends up with Kerry having the higher % chance of making it while they end up with the slightly higher Latest xP.

Huge fan of the S&S twitter account. They're producing some genuinely fantastic information this year. Not getting nearly the attention they deserve yet. Hope it bumps in popularity to give them plenty of incentive to keep it going.

The FAI connect app is genuinely revolutionary. Mayo soccer socials had been extremely lacking for some time (check out some club's Facebook or Twitter pages and their last posts are from 2017). Even with things like upcoming fixtures, team lineups for matches etc. there was such a void of information from Mayo soccer clubs online, save for maybe Kenneth Gannon's very good work for Westport Utd.

Nearly overnight that's been turned around with the new app. Would be a game changer for GAA clubs if something similar could be developed. I don't think it would be too difficult logistically either, as the FAI connect app is fed by the database that the Mayo League has regarding goals, assists, cards etc.

Can't argue that the soccer has been lacking but they've been trying / making an effort for the last 5 years or so. I was managing a team 2018 -2020 and they used to have a weekly video round up / interviews which created a good buzz and was probably the start of it. But some clubs just didn't have the people interested to keep their social media updated.

Kenneth has been doing a great job for a long long time now. Celtic have this new Veo system for live streams which is also very impressive. Ballyheane had the first ever Mayo League game under lights last week, another positive step.

There's a few good young lads on the Mayo League committee now trying to drive things on. The weekly podcast is very detailed and a credit to them.

Something like FAI Connect would be excellent for GAA, shouldn't be that hard to do, my comment was more that it would be a harder job to keep it accurate with all the scores, various cards etc for a GAA match.

You'd be the only one outside of Charlestown hoping for that Par

Hi folks,

Someone on Reddit directed me towards some discussion about my model on here so I've just popped on to explain a bit about it.

The model works a bit differently to others I've seen which typically use some form of the ELO method (winner takes points off loser). This is basically because it was a bit of a project to see could I create something "decent" purely from scratch.

Roughly it works by taking every game and saying "what would this team have scored and conceded vs a hypothetical average opponent in a neutral ground in the middle of July 2022" and then it considers a range of recent performances to get to a rating. It does this for goals and points separately.

This caused me all sorts of headaches and it has as many as if not more problems than other methods, just different ones! One thing I like about it though is wins can be viewed negatively and losses positively, which I think reflects reality. Home advantage can also be quantified explicitly by team.

On the %s to finish in each spot, each game has a predicted scoreline which I then treat as an average and do some maths-ing to simulate 1000 possible outcomes based on that average. So my model will have Kerry beating Roscommon in its prediction but in the 1000 scenarios they may win 85% of the time. As TsuDohNimh has already explained the xP vs % finish apparent contradiction comes from ties. Because the season is so short AND it is only 2 points for a win there are just so many instances of teams ending up on the same number of points. And therefore it can be the case that a h2h advantage (whether that be directly or through PF in a 3-way tie) can be worth a little more than a small amount of expected Points.

Specifically with Division 1 I would say the model is assuming Kerry will perform in their next 2 games at a level they have been nowhere near so far. My model is very slow to change its opinion (depressingly in most circumstances that seems to be the best approach). I think in the case of the game v Galway at least it is fair to say they are less likely to win as many of the 1000 games as the model will be forecasting (it has Kerry by FOUR). This not only means their xP is a little high but Roscommon & Galway's is a little low.

Dempsey gone from the Limerick job....sounds like he resigned but it's fairly weird timing either way.