Chapter Tactics is a 40k podcast which focuses on promoting better tactical play and situational awareness across all variations of the game. Today Peteypab is back with your regularly scheduled Chapter Tactics. Tournament coverage and analysis of stats from tournaments pre- and post- the big September FAQ.
Show Notes:
- Want to check out the stats we viewed in the episode? Click on this link.
- Check out The Honest Wargamer Rob for more AoS and 40k coverage and tactics.
- Don’t forget to check out our new sponsor! Broken Egg Games, and Rum Runner Wargame Painting and Conversions.
- Click here for a link for information on downloading best coast pairings app where you can find lists for most of the events I mention.
- Check out the last episode of Chapter Tactics here. Or, click here for a link to a full archive of all of my episodes.
- Commercial music by Music by: www.bensound.com
- Intro by: Justin Mahar
Need help with a list idea? Got a rules question? Want to talk tactics? Then email me at…
frontlinegamingpeteypab@gmail.com
Please do not send an army list in a format such as Army Builder, send them in an easy to read, typed format. Thanks!
Thanks for the shoutout Val <3
Interesting topic – I find the discussing and theorizing into why and how lists/armies are being taken compelling. Especially when paired with player personality etc. The depth of statistic available is growing and while I think we are in the infancy stage of actually analyzing the data – the guessing and speculating is fascinating on its own.
That was pretty cheeky Petey Pab saying that Necrons need to start using Nihilakh Dynasty.
For those that don’t get it that is the “if you don’t move reroll 1’s for shooting” one and for an army that is 99% a 24″ range army there is only one list that you take Nihilakh as your primary…. Vault Spam for the Strat alone. 😛
I am almost positive that GW will break Vault Spam in CA as they have done will all one dementional gimmicks like Flyer with Gman Spam and Flyrants as will. I am looking forward to it as I think it is a lame army to play and play against.
I’ll be “that guy” and complain about the methodology, but it’s not the sample size or what have you that I want to mention. I just really think you need to specify your confidence level if you’re going to talk about things at the 3-5% level. Is a 45% win ratio statistically significant based on the data? What are the confidence intervals like? I realize you need to specify an underlying distribution of some sort in order to do this in a reasonable way, but assuming player skill is normally distributed is probably a fair assumption to make.
Maybe just let the null hypothesis be that the “real” win ratio is 50/50 and compare to a T-distribution to see what the odds are of getting the results by chance. I’m not asking for p = 0.05 but an idea of the significance of the results I think is really important.
Thanks, you are correct and I have been working on a +- calculation that I would feel comfortable assuming using the current data. There are a lot of variables at play though and I actually don’t feel that player skill is evenly distributed cross-faction of yet. I believe I mentioned the first time I was on the podcast that I felt a 3% deviation was not significant enough and should have brought it up again here.
Hey man, I’m sorry and I don’t mean to nitpick or anything, I just deal with numbers and statistics and means and variances and what not all day, and my first question upon hearing a statistic is to ask “at what confidence level?” (Or “what’s your alpha level” or what have you).
I was actually trying to brainstorm a way to use Bayesian statistics so that you could assume a normal distribution with mean 0.5 as your prior and then your results would converge to the true mean the more data you got, but it makes getting confidence intervals really hard so it’s probably not a worthwhile exercise.
I think in a previous episode you tried to control for player skill by comparing the same person’s win ratio when they played different factions, and that’s probably the ideal way to do it but it requires a lot of data from people who have played multiple factions to be reasonable. Another way to do it might be to compare their win ratio in mirror matches, ie how often do they win against their own faction, and then normalize to that to correct for player skill. Probably only practical for really popular factions like knights and ynnari though? I doubt there are too many Thousand Sons mirror matches to get decent statistics from.
I hate making these suggestions without having the time to actually do the calculations myself, it comes off as “hey why don’t you do a lot of work for my amusement.” That’s not my intention, I’m just thinking aloud (can you say ‘aloud’ when you’re typing? Anyway, you know what I mean.)
Hey guys, do you know where I would be able to find the TIWP for the different factions, or how to calculate it with the data?
Hi guys,
Great episode, and I totally agree that we need more stats for the game to get a real idea of how factions are performing, both in absolute terms and against each other.
One thing I will mention is that your percentages are averages of each individual win percentage, rather than a win percentage of the data set as a whole.
This means that each tournament performance is weighted equally when calculating the overall win percentage, even though some tournaments have more rounds than others.
In most cases, this doesn’t have a huge effect on the final figure, but the lower the total number of games, the more this has an impact on the overall win percentage. The downside of this is that there’s no easy way to calculate the overall win percentage within a pivot table, so from a practical point of view it might not be the best option.
Just my 2p worth!
Thanks for the feedback. I always considered the win % I was calculating as simply an ‘average tournament win %’ rather than overall win%. However, I do have the data and can easily create a ‘weighted’ win percentage column in the pivot table.
That’s a great point and one I hadn’t actually considered. I think I had just assumed it was supposed to be an overall win percentage, as this is a common mistake for this type of calculation! Thanks for clarifying.
Hey everyone great feedback, keep it coming!
Exactly the sort of thing I listen to Chapter Tactics for. It’s been a rocky month or two but this is the good stuff.
Also you should see the lack of comments as a positive sign. Negativity seems to get people talking much more, and when there’s a bunch of comments in this section its always, without fail, either a troll in the comment section, or serious complaints about the episode. This one hasn’t triggered either, and gotten a few positive comments.
also, can we please get a link to the numbers?
Link is in the show notes.