Excel LADZ
patreon
Excel LADZ - May 2024 Update
🕑 Added 2024-05-20 02:19:59 +0000 UTCG’day, lads!
Football Model
Thank you all for watching my most recent video, ‘ANALYSE Football Matches for Win %, Over/Under and Correct Score Using EXCEL! | Tutorial’. The video has reached over 2,000 views in its first two weeks; a new channel record!
Download this model here: https://www.patreon.com/posts/analyse-football-103833026
With the EPL Season done, and most major European competitions ending soon, Patrons have been requesting that Summertime competitions (e.g. MLS, J-League, etc) be added to the model. This is entirely possible, and will be done very soon! I’ll record myself adding these leagues so that it’s possible for people to add their own competitions to the model.
MLB Model
With the MLB Season in full swing now, it’s time to update the ‘PREDICT MLB Matches with Excel | Tutorial’ YouTube video. This tutorial builds an entirely team-based model that calculates an Expected Runs figure, and then uses a custom distribution to find the Win % for each team.
The update to this model, which I’m planning on being my next video, will incorporate Starting Pitcher stats and Team Lineups to make the model’s forecasts more accurate. Batting and Pitching Stats will act as multiples for a team’s xRuns. For example, a pitcher that has performed 1.1 times better than his team’s average will decrease the opposition’s xRuns by 10%.
Of course, these stats will be brought into Excel using a Power Query so that they can be live and updated. If you have any questions/suggestions for this model lads, please leave a comment on this post!
AFL Disposals Model
For my Australian followers, I’ve just created a model to forecast a player’s disposals in a match. Using a binomial distribution, a player’s xDisposals (calculated by considering their average, adjusted for the strength of their opponent) is simulated to find the probability of them having over/under a particular threshold. For example, Pendlebury Over/Under 20.5 disposals can be found. The download for this model is attached to this post.
An NRL Model is coming soon!
Thanks for reading lads! If you’re interested in sports modelling, then be sure to join the Excel LADZ community!
Comments
Randy Reed
Last night is a great example. Brewers starting pitcher Joe Ross was pulled about 30 minutes to 1 hour before the game (back strain) so the Brewers brought up a pitcher from triple A and inserted him into the starting pitcher slot, and then relied on bullpen for the playable pitchers. The percentage for the Brewers to win dropped due to this, but with most models you can not catch this if it happens. The ability to be able to check starting pitchers and starting lineups as they are released will help to determine the best odds near game time.
Excel LADZ
The community is great 🔥 I very much understand your problem, and that's too much frustrating manual work to have to do. I try and approach this problem by having my player data centralised. So all player stats come from the same source, meaning player names don't have to be reconciled. This is obviously difficult, and is all about deciding the best data source. On the other hand, there are only so many teams in a competition. That's when you can afford to use different data sources, when pulling in team stats, as you can just match different team names (e.g. Lakers and LAK) with a simple XLOOKUP on a created table. With the MLB Model I'm going to make using Pitcher stats and daily lineups, I'm going to use player stats to come up with multiples to apply to the team's Attack and Defence Ratings. For example, a relatively good starting pitcher for that team will boost their ATT, and thus their xRuns. I hope that helps lad. Of course, anymore issues you have you can pop it in the Discord 💪
J. Deisch
I agree with Randy. The most difficult issue I've had in trying and failing at my own model is the sheer number of name reconciliations that must be performed across each data set across all the variations of names. I am so thankful for a community of LADZ that supports each other on our journey.
Randy Reed
Like I had mentioned earlier, Pitcher Stats, Bullpen stats, Team lineup stats........ when daily lineups are released cause that will affect the outcome of a game as well. Bullpen stats help a lot too because starting pitchers most of the time now go around 100 pitches and then they go to the bullpen. Some calculate the pitcher/bullpen ration around 60/40, if that makes sense. Need to be able to see the pitchers ERA for each team playing, and also team hitting avg if that's possible. I think if you incorporate these stats the model will be a lot more accurate!
