Hi everyone. One of my roles on the Steering Committee is managing our game data collection and analysis so we have an objective, big-picture view of how cards are performing in the IACP. It’s a common saying in business that you can’t manage what you don’t measure. While we can’t collect data on every skirmish game that ever gets played in the IACP, nor can we ever collect enough data to have a sample-size large enough to give us a perfectly accurate view of the meta, we like having at least some quantified data to look at and compare with the feedback we get from the community and with our own personal experiences and evaluations. This is something I’ve been working on since the Season 3 Approved season, and something that I’ve been getting better and better at the more I learn about game data analysis and Google sheets formulas. The IA Community has been a huge help with this effort by submitting vassal game logs to the IACP Dropbox to give us the data that we need.
With over 100 Season 5 playtest games submitted and processed through our IACP dropbox system, we’re starting to get a decent look at how the Season 5.0 changes and new cards are actually doing on the field of battle. This data represents all of the games that were submitted to us before March 5th. I felt that 101 games was a good stopping point to take stock and do some analysis for you all, but be assured that games submitted from March 5th through March 8th will still used to quantify cards that did not get a functional change on March 8th, so we’re not “wasting” any of that data just because it happened on a lame-duck weekend.
Let’s start off by taking a big-picture look at the Season 5.0 meta, and then we’ll look at the Season 5 cards specifically, and also go into isolating some independent variables to clear up some confusing spots in the data. Feel free to look at the data in full here if you want to see more information (such as performance by faction).
This shows you every deployment card that has seen play in the playtest logs we’ve gotten so far. And gives you a very zoomed out, deconstructed view of what the meta has looked like so far. It’s important to remember that the smaller the number of games played for a card, the less accurate that card’s win-rate is likely to be, because even just a few more game results could significantly increase or decrease that win-rate % number. We don’t like to use win-rates of cards with less than 15 games played as justification for making changes, though we do use those numbers as a reason to investigate further and push for more getting more quantitative and qualitative data on those cards. So if you see a card that has a very high or very low win-rate, it’s important to also check how big that number in the games played box is before you panic or pick up your torch and pitchfork.
For that reason, you’ll notice that a lot of the cards with less than 15 games played have very high or very low win rates (orange or cyan colored), whereas cards with more than 15 games tend to more often have settled into that nice green color, between 40% and 65% win rate.
Oh and of course we also have data on Command cards as well.
Season 5 is the first time that we have tried to measure Command card usage and win rate. This was always a challenge previously because when people submit Vassal logs, there’s no way to view hidden cards that don’t get revealed, like Command cards that don’t get drawn. And unlike deployment cards and skirmish upgrades that start the game in play and always have at least some impact on the game, the random nature of the Command deck means not every Command card gets played for its effect, so we (I) have to also carefully examine each game log to see which Command cards were played and which ones were either discarded or never got drawn. There are 2 different metrics for Command cards, popularity and effectiveness. Popularity, or meta share, tells us how many people are spending card slots to include that card in their deck, even if they never get to play it. Effectiveness tells us the correlation between a card being successfully used for its effect and how often its connects to a win for that player. Just like measuring win rates for deployment cards, we can’t say for a fact that a win or loss was caused by one card resolving, but because we know that game results are at least partly determined by the effect of the cards being played, we can say there is a strong correlational relationship between a card resolving and the win rate for that cards player.
Unfortunately I’m not able to easily count which Command cards have mirror matches the way I can for Deployment cards yet because the very long formula I used to measure mirror matches for deployment cards will count unplayed ccs as well, which is not what we want to do, so for now if we need to adjust for a mirror match to get a more accurate win rate, we’ll be doing it manually. It’s important to remember that ultimately mirror matches just skew a card’s win rate closer to 50%, so including mirror matches won’t ever be the reason that a card has an especially high or especially low win rate.
Before we dive into the Season 5 IACP cards themselves, I’d like to discuss a few cards that aren’t from Season 5 whose win rates might look a bit alarming. Something that I’ve started doing more of in Season 5 is to isolate different variables from one another to make sure that certain data points aren’t being skewed by certain variables that don’t accurately reflect a card’s performance under normal or optimal conditions. This is also a good way to figure out why old cards suddenly have much higher win rates than they should based on previous performances.
Black Market: 92% is very high for a card with 19 games played. It’s also a huge jump from the data we had collected in Season 4 on the card, which showed it at a 55% win rate from 51 games recorded. I had theorized that Aphra was a big contributer to this jump because of her ability to reuse powerful VP Command cards like Celebration and Black Market Prices, as well as her Command card, all of which is now much less powerful in 5.1. However, Black Market’s win rate with Aphra in 5.0 was 80%, compared to 77.78% without Aphra, which doesn’t signify her as the big contributor to Black Market’s high win rate. Mara Jade also had a Command card that involved gaining VPs, but did not significantly boost Black Market’s win rate either. Jabba certainly improved Black Market’s win rate, but not above 90%, and that pair had already existed in previous seasons. Upon further investigation, HK-47 was actually the biggest skew on Black Market, with 100% of games won when the 2 cards were combined, while Black Market’s win rate dropped to a much happier 60% without the murder droid. We’ll talk about why HK-47 wasn’t actually the powerhouse card some people thought he was in further down, so it may just be a case of a strong card boosting a relatively weaker one, but currently I admit I don’t quite understand the nature of what makes this combination so strong and why it has such a warping effect on Black Market’s win rate, but this is something I will continue to monitor now that HK-47 has gotten a change in 5.1.
Gaarkhan: 75% seems very high to me for this card, and is just very high in general and warrants investigation. In season 3 Gaarkhan had a 47% win rate, and in Season 4 we did see him jump to 61%, but in both seasons we had barely more than 20 games recorded for the figure. The obvious theory would be that Fury of Kashyyyk was boosting his numbers in Season 5, but Gaarkhan has actually won every game without FoK and only wins about half the time with FoK. The Season 5 deployment card that seems to have the most positive influence on Gaarkhan’s win rate seems to be Mara Jade, who increases Gaarkhan’s win rate from 60% to 87.5%, but Mara’s win rate is not particularly amazing on its own, and pairing Mara Jade with other figures actually seems to lower their win rate as often as it increases or stays the same. Something similar happens when Mara is paired with Jedi Luke, who has a similar Command card effect, raising his win rate from 50% without Mara to 83% with Mara, though with only 8 games played with Jedi Luke, that data is not particularly reliable. The obvious answer here would seem to be Furious Charge, except that the data shows that only 1 game was recorded where Furious Charge was successfully played in a list with both Gaarkhan and Mara, and we don’t see something similar with Mara combining with figures like Chewbacca and IG-88 that have ccs with similar effects. This one is also a bit of a riddle to me, but we’ll continue to monitor Gaarkhan as the playtest continues and see if things don’t even out after the 5.1 update or if perhaps we need to take action.
Season 5 card stats
Here we can look at just the performance of the cards from Season 5. This is where I’ll talk about the changes that were introduced for 5.1, because looking at this data there are some oddities that might stand out to people in comparison to the 5.1 updates.
Let’s start with Scum. The highest win rate of any season 5 card with more than 15 games played, the elite Clawdite stood out as doing exceptionally well with a 75% win-rate. This was much higher than the regular Clawdite’s 48%, which told us that the elite’s better health and attack power was an important factor. While the Senator, Scout and Streetrat were the most commonly seen forms, in lists that brought 2 elite Clawdites, which upped their Spy count without giving up any Hunter or Rebel slots, the soldier’s strong ranged attack profile was used to complement the Scout’s to allow both clawdite figures to be able to function as strong ranged attackers on the same turns.
In addition to all of the feedback we were getting about the oppressiveness of Under Duress milling 2 cards each round on top of the base effect, it was also doing very well in the win rate department, 64%. For HK-47, we actually weren’t seeing a especially high win rate, and in fact the vast majority of HK-47s were in combination with Under Duress, where the combination of [Query] and Under Duress were obviously powerful. Since we knew we were going to be nerfing Under Duress, we decided that we would try to keep HK-47’s power level mostly the same while adjusting his abilities to make them less swingy and less frustrating to play against.
Doctor Aphra looks like she should be quite balanced, but here the data is hiding an important quality of Aphra’s, which is she has 2 modes. One is when she is played with 0-0-0 and BT-1, and when she is played without the droids. While a lot of people at the beginning of the season predicted that Aphra would need to be played with her droid counterparts to get her full value, the exact opposite actually proved to be true. Aphra only wins 11% of her games when played with 0-0-0 and BT-1, and wins a whopping 70% of games without her droids. Which showed us that Dubious Counterparts was not strong enough to overcome the addition of 0-0-0 to a list, and also that Excavation was quite a bit more powerful than we initially hoped. Indeed, the shenanigans that were made possible by Excavation, like emptying an opponent’s hand on round 1 with 1 copy of Hostile Negotiations, or hitting 2 figures for 6 strain on round 3 with 1 copy of Running out of Time proved to be too much, which is why we changed the card to not be able to play the same card twice in a round, and to no longer generate Command card advantage.
The Wampa was one of the most popular Season 5 figures in the first 2 weeks of the league, but it constantly struggled to get game wins and consistently languished with a win-rate % in the 20s and 30s, despite a majority of players utilizing the obvious combos of Beast Tamer and Pummel or Wild Fury, along with Apex Predator, Dying Lunge, Ferocity and Parting Blow. No matter how we isolated the data to try and see if there was a “bad” card that players were using that was bringing the Wampa’s win rate lower than it should be, it never brought it above 40%, nor did the “good” cards like Extra Armor or Beast Tamer.
In Rebels, Murne was another card that seemed to have a very balanced record and also fit into a ton of different lists, but it was easy to see from even a few game logs that when there were good targets for it in the opponent’s list, an actionless False Orders was very difficult and frustrating to play against.
Much like the Wampa, K-2SO has seen a lot of play in Season 5 but has put up surprisingly poor results, and there doesn’t seem to be an obvious culprit figure that is bringing K-2s stats down. In fact, K-2 seems to have a chilling effect on any card that he gets paired with, and is responsible for bringing down the win rates of quite a few other rebel figures, including Davith and Cassian’s, which is why we knew we didn’t need to buff those cards despite their lower win rates at first glance. In fact Davith is looking quite powerful when you factor out the games that he was played next to K-2, with a 57% win-rate.
Fury of Kashyyyk’s win rate is also decent but less than 50% despite getting a nerf in 5.1, but that was because we were getting a lot of feedback from players in the league that claimed the Wookiee list was frustrating to play against because of it’s high amounts of Health and ability to hit hard. Looking deeper also reveals that factoring out games played with Chewbacca in the list raises Fury of Kashyyyk’s win rate to a more impressive 60%, which shows that FoK list that sticks to Wookiee Warriors and Gaarkhan seems to be quite stronger than the base win-rate would suggest.
Empire was the most balanced faction for season 5 cards, with the only stand-outs being, surprisingly, the Heavy Stormtroopers, which not a lot of players seemed to be that impressed by at 8 points. We’ll continue to monitor the Heavies since their win-rate is high but has not quite stabilized yet with only 12 games played. The most common attachment with the Heavies was the General’s ranks, but sadly the vanilla Targeting Computer had the highest win rate with Heavies.
Kallus is doing just fine at 50%. Iden might look like she’s in trouble at 35%, but when you factor out games where she was played without Zillo Technique, her win-rate jumps to a healthy 45%. Terro’s win-rate goes up to 50% when factoring out 3 games played with Scout Troopers, which are well-known for being fairly underpowered in the current IACP meta (and imo, every meta they’ve been in). The Generals ranks is doing nicely with a 53% win rate. Mostly we’ve been seeing it used on Heavy Stormtroopers and ISB Infiltrators as expected, but it’s also started to see play on AT-DPs and the AT-ST where it’s quite effective at super-charging their movement from Officers to get them into position for a nasty barrage with Assault.
Finally for the Deployment cards, Mara Jade has been extremely popular, but has not performed in testing the way a lot of people had been worried she would, and her 53% win-rate reflects a solid but fairly balanced figure. While her attack does a decent chunk of damage, her relatively low health makes it hard for her to last very long once she’s engaged to make her attack unless you . We haven’t really seen any of the crazy combos that some people have theorized about actually play out in gameplay, though the combination of Adaptive Skills and Fast Learner does give her a ton of flexibility in being able to pretty much always have a good Command card that she can play, but she doesn’t actually do anything special outside of the Command cards that get played with her. Who she gets played with actually seems to have more to do with her traits than with potential Fast Learner combos. She’s been a big boost the The Grand Inquisitor for example as a fellow Force User Hunter.
Looking at the Command cards, we see a lot of the same things happening here that was already discussed above. Apex Predator and Blend In likely had their win-rates affected by the low win-rates of the Wampa and K-2SO. Definition: Love’s was affected by the high win rate of K-2SO with Under Duress. Interestingly, Overwhelming Impact had a much lower win-rate than I think a lot of people were predicting, below 50%. And the 2 games where Overwhelming Impact and Assassinate were resolved by the same player in the same game resulted in losses, which is also true of when Tools and Heightened were played alongside Overwhelming Impact, so it seems the prophesized combo of Overwhelming Impact + Hunter cards with BT-1 has not come to fruition. Inversely, Feint seems to be overperforming based on the community’s initial reaction, but 9 games isn’t a lot and the win-rate will likely come down as more games are played. The same is true for everything else that had less than even 10 games played, we’ll have to either wait for more data to come in or go by what we see in game logs and player feedback with these cards.
And then there’s the unnerfed Assassinate, which was not only the most played Command card from Season 5 by far, it was also the 10th most played Command card out of the 168 Command cards that were played in the league. While it’s win-rate is high at 60%, there are only a couple of Command cards with similarly high levels of play that rival it, such as Deathblow and Dying Lunge (the strain cards with 70+% win-rates are of course the result of Under Duress’s warping power over the 5.0 metagame and should be discounted). Since 60% puts Assassinate in the higher win-rate levels but not so high above some other cards that are not being tested for a nerf, we’ll be gathering more data on Assassinate in its unnerfed form.
So that’s pretty much the data we got from the first part of the Season 5 playtest as I’ve interpreted it. Hopefully this helps put all of the data into context for people. It’s also important to remember that we’re looking at discussions and feedback on these cards and not just raw data. It’s really important to always try to get a holistic view on things so that we’re not missing the forest for the trees. Make sure to let us know what you’re experience with the Season 5 cards has been like so far, either in the comments below, in the Season 5 playtest survey, or on the Zions Finest Slack and our other social media outlets.