Dominion Strategy Forum

Please login or register.

Login with username, password and session length
Pages: 1 [2]  All

Author Topic: Data Mining: Card Impact Factor  (Read 1294 times)

0 Members and 1 Guest are viewing this topic.

markus

  • Conspirator
  • ****
  • Offline Offline
  • Posts: 210
  • Shuffle iT Username: markus
  • Respect: +218
    • View Profile
Re: Data Mining: Card Impact Factor
« Reply #25 on: May 01, 2018, 03:22:03 pm »
0

Or like, on board where you go for a Stash deck and you buy Duchy, Duchess is a "might as well" addition. Or am I misinterpreting the statistic.

It's the other way around: on a board with Duchess, you're more likely to go for a Stash deck. (We don't know the number for Duchess on a Stash board.)
Logged

markusin

  • Cartographer
  • *****
  • Offline Offline
  • Posts: 3639
  • Shuffle iT Username: markusin
  • I also switched from Starcraft
  • Respect: +2202
    • View Profile
Re: Data Mining: Card Impact Factor
« Reply #26 on: May 01, 2018, 04:07:29 pm »
0

Or like, on board where you go for a Stash deck and you buy Duchy, Duchess is a "might as well" addition. Or am I misinterpreting the statistic.

It's the other way around: on a board with Duchess, you're more likely to go for a Stash deck. (We don't know the number for Duchess on a Stash board.)

But there's overlap isn't there? Like, Masterpiece/Feodum has a similarly high synergy factor to Feodum/Masterpiece.
Logged

markus

  • Conspirator
  • ****
  • Offline Offline
  • Posts: 210
  • Shuffle iT Username: markus
  • Respect: +218
    • View Profile
Re: Data Mining: Card Impact Factor
« Reply #27 on: May 01, 2018, 05:13:36 pm »
0

Or like, on board where you go for a Stash deck and you buy Duchy, Duchess is a "might as well" addition. Or am I misinterpreting the statistic.

It's the other way around: on a board with Duchess, you're more likely to go for a Stash deck. (We don't know the number for Duchess on a Stash board.)

But there's overlap isn't there? Like, Masterpiece/Feodum has a similarly high synergy factor to Feodum/Masterpiece.
Yes, you'd usually expect that. For Feodum/Masterpiece both cards are not getting gained much without the other, but gained a lot when they are both present. For Capital/Mandarin I'd expect that Capital increases gain% of Mandarin more than the other way around.
Logged

trivialknot

  • Minion
  • *****
  • Offline Offline
  • Posts: 588
  • Respect: +919
    • View Profile
Re: Data Mining: Card Impact Factor
« Reply #28 on: May 01, 2018, 05:18:41 pm »
0

*ok, I have no idea why Duchess promotes Stash.

I do! When you gain a bunch of terminals for free, you don't have so much terminal space left so you'll buy Stash or Venture instead of a terminal $5. I don't think this effect would be at all visible today, but I'm like 90-93% sure that was really the reason in 2014.
I guess that makes sense.

FWIW there's less data for the promo cards, so they tend to have larger error margins.  That's why the promo cards show up disproportionately in the top synergies lists (and also why I had to remove Prince/Walled Village).
Logged

markus

  • Conspirator
  • ****
  • Offline Offline
  • Posts: 210
  • Shuffle iT Username: markus
  • Respect: +218
    • View Profile
Re: Data Mining: Card Impact Factor
« Reply #29 on: May 28, 2018, 05:50:00 pm »
+2

Thanks to ceviri and his tool, I got some logs from the new client since the Nocturne release. There are about 1100 logs that feature either me (75%) or RTT as a player.

The small sample means that you can expect some outliers for random reasons, but the results look fine enough to me that I'm posting them. You can find them here.

Small sample also means that I'm calculating the impact factor a bit differently than trivialknot, but the idea is still to see how much the buy% or gain% of other cards changes in absolute terms when card/event/landmark X is on the board. Results are normalized to have mean 0 and standard deviation 1.

One thing to note is that a card being on the bottom means that it doesn't affect the chance that at least one of the players buys/gains the other cards much. That could be, because it's a dud as for many cards in the original list. Or because it doesn't affect the average strategy much.
Probably, Herald is at the bottom for random reasons. But you could make a case that with Herald on the board you often play it the same as "on average" and you gain some Heralds in addition.

PS: There's a bunch of other stuff, I've been calculating with those logs, which you can find on the other sheets. Feel free to ask, if there are any questions.
Logged

Commodore Chuckles

  • Duke
  • *****
  • Offline Offline
  • Posts: 373
  • Shuffle iT Username: Commodore Chuckles
  • Respect: +631
    • View Profile
Re: Data Mining: Card Impact Factor
« Reply #30 on: May 28, 2018, 06:13:09 pm »
0

Yeah, Herald is basically a helper card that usually doesn't encourage you to buy things you weren't going to buy anyway. I don't think it's a coincidence that Lost City, which is basically a Herald that never duds, is also near the bottom. Though Scheme is near the top, and that's also just a helper card, so... dunno.

One thing I am surprised by is how low Stonemason is, considering the way it changes how you gain stuff, like Alms, which is near the top.
Logged

markus

  • Conspirator
  • ****
  • Offline Offline
  • Posts: 210
  • Shuffle iT Username: markus
  • Respect: +218
    • View Profile
Re: Data Mining: Card Impact Factor
« Reply #31 on: May 28, 2018, 06:41:55 pm »
0

Similar to Alms, Stonemason also has the property that it changes gains less than buys. It was below average on the list for gains with cards up to Guilds as well.
Logged
Pages: 1 [2]  All
 

Page created in 0.245 seconds with 20 queries.