Thanks to those who made this post possible: I would like to thank two people, without whom this post would not be possible. First is DB0, the creator of the OCTGN Netrunner plugin. Without him we might not even have online Netrunner, and we also wouldn’t have the OCTGN game results dataset, which is massively valuable to anyone trying to gain a good understanding of the relative strength of the different identities, based on actual data. Several weeks ago, DB0 released a set of data for over 150,000 Netrunner games, which is the basis of this post. Second is Captain_Frisk…
not much else to say other than the stats seem to reflect my live play experience. i gauge the power of the ids by how much i/ my opponent roll their eyes when they see a particular id across the table.
andy/nbn draw the most eye rolls and scowls of disgust by far right now
my least favorite id to play against is actually personal evolution because you have to super methodical/ disciplined to win and sometimes im just not in the mood to play that way.
I always thought Noise was a higher skill deck and puzzled when Gabe players claimed otherwise
It was charitable of you to say that both Andy and Gabe advocates got something right and wrong, but I thought what the debate cared about was performance at high level of play, of which Andy is stronger (now set on even firmer grounds by this data analysis)
Most of us Andy supporters thought she was really 6% better on average. But its actually more like probably 2%, if a true mix of all players were to play both Gabe and Andy.
On Gabe and Kit gaining the least from player skill, they’re both the “Aggro” IDs for their respective factions, and I think the challenge falls on the corp rather than the runner in those games (“Can you keep out Gabe/Kit, if yes enjoy the win”).
Intresting article, great job!
But to be honest I’m not sure in definition of strong player. I’ m very intresting in similar analisys with 60% or even 65-70% cut. As far as I remember Gabe-guys claimed that he is the best at very high player level, not at over-average-skill
I really enjoyed reading the article. I wonder if it’s possible to record decklists in the future in order to drill down into archetypes within an identity. While for some identities it’s pretty clear what the best build is, for some of the Tier 2 identities I think the jury is still out.
For example my personal experience is that for Kit, Cyber-Cypher builds are significantly stronger than Yogosaurus builds, though both are worse than Andromeda - it would be really useful to have data either way on this.
Yes, this was the claim.
I’ve seen an older analysis that gave players a post-hoc ELO rating, which facilitates comparison at the highest level.
Elsewhere I’ve also seen an analysis that considers the Corp’s agenda total when winning by flatline. This was in an attempt to defend Jinteki by saying that they force the runner into riskier plays at match point - but actually the data showed the opposite and Jinteki won more in the early game by the runner dying stupidly when there’s no threat.
I think that if you want to really drill down on very high skill play, you need to use ELO ratings, not winrate.
Putting a bar at 60% or 65% or whatever could cut out some otherwise excellent players who mostly just play games with other top players.
It would be important to figure out ELO and then cut the top X% of players by ELO, and then look at those games.
Regarding my ‘strong’ player set not being good enough for the Gabe claim. Its true that this cannot definitively answer the question, in that it doesnt have stats for just specifically the very elite players. However, the percentages are trending in favor of Andromeda as you go up in skill, so I do think that is leading you in a certain direction.
@Alexfrog, I’m also interested as to whether you applied any cuts to your data set?
i.e. discount any games with abnormal amounts of influence (to remove illegal decks, or beginners using the core decks); turn 1 runner wins (since these are entirely luck based and not a true reflection of skill or ID performance).
You also say you discounted Fisk and The Collective, but what exactly do you mean by that? Do you mean that you just disregarded them for the purposes of analysis or did you actually discount those matches? The Collective will have contributed to a significant number of Corp losses. Similarly it will have biased your determination of “strong” and “weak” players - since some mid-range but good Corp players will have lost more than half, and mid-range bad runners won more than half due to a broken ID.
I think if you went back and applied a post hoc ELO (or rating system of choice) you could have a bigger dataset, because you wouldn’t have to restrict yourself to people who have played 20+ games. If someone only plays 5 games vs. high ELO players and wins them all that’s still a reasonable indication of skill.
You could then look at ID win-rates in 5% bands of skill. The data set would definitely then be smaller, but it would be a start. We’ll never have a large enough dataset in a stable meta to make any firm conclusions about anything.
Possibly speak to David Sutcliffe (Magicdave on BGG). He’s very good at stats and probably already has some scripts written from his last analysis.
hmm, does this data set go back to the beginning of OCTGN? i assume if you broke it into pre and post jackson howard era you’d notice some differences, especially in noise.
regardless, nice article. your comment about noise and hb being more rewarding to stronger players makes me feel even better about my gencon finish
@Alexfrog are you able to query the OCTGN data to check the prevalence of certain cards?
I’ve been having a disagreement with a guy on BGG over the effectiveness of traps (he thinks they’re great, I don’t) and I think it would be interesting to see if there’s a discrepency in how often they’re played in different Corp skill bands.
Looks like the whole “Spin Cycle is pro-Corp” rumor can be pretty thoroughly confirmed at this stage-- Corp win rates have gone up with every pack. It’ll be interesting to see if this trend continues until the end of the cycle.
