The Plugged in Tour, Worlds, Andromeda - A Brief Look at Data

The best players in my environment definitely think Gabe is better at a high level of play. It’s possible that my environment is bad, but that doesn’t seem particularly likely to me at this time.

To believers in ‘Gabe is better at high skill’, what would it take to disprove your hypothesis? What would have to happen in order for you to change your mind?

If Gabe had been 7/8 of top 8, 13/16 of top 16 at Worlds, and no Andromeda in top 32, I would most DEFINITELY change my belief and say: “Okay, wow, Gabe was better!”

If Gabe had been 6 of top 16 and Andromeda 6 of top 16, and less people played Gabe, I would say: “Okay, thats some good evidence that Gabe actually performed better at the top level in worlds. I’m going to strongly consider that”.

If Gabe had more tourney wins, or higher win rate in a large data sample, or something, I’d give it a lot of weight. Just like when NBN:TWIY did well on OCTGN, it changed my opinion on it.

What actually needs to occur in order to make you think Gabe isnt stronger? To me, the worlds result was pretty much the strongest possible evidence in favor of Andromeda>Gabe that I could ever realistically see. I mean, THIRTEEN OF TOP SIXTEEN. Holy crap! NO GABES. AT ALL.

1 Like

We would need to have access to information regarding the base rates of Gabe and Andromeda play. For instance, let’s say Gabe actually outnumbered Andromeda at Worlds. I would consider that very strong evidence in favor of Andromeda being better than Gabe!

It’s pretty difficult for me to weigh-in here without coming across like a troll, but I’m going to anyway and make a few, I think pretty relevant, points.

  1. If Gabe was equally as good as Andromeda (not saying he is, just assuming for a second) then you would expect them to be proportionally represented at the business end of tournaments - i.e. if there were 9 Andromedas to every 1 Gabe then you would expect that distribution in results as a first order approximation.
  2. However Kingsley’s assertion is that Gabe is favoured by elite players. If that were really true then you might expect to see Gabe over-represented in the finishing positions - it stands to reason that if the better players are driving him then they’d do slightly better than the “weaker” Andy proponents.

That’s not the picture we’re seeing, in fact it’s quite the opposite - Gabe is under-represented. Conclusion: One (or both) of the assumptions is probably false, so either Gabe is weaker than Andy and/or the players playing him are worse.

Now I’m not suggesting that tournament wins tell the whole story. Variance is a cruel mistress and any deck can beat any other on a given day - it could be the case that every runner-up on the PIT was a Gabe deck that just got unlucky. However, he’s not even making the top 32, while Andy is dominating the field.

Actually that seems very likely to me at this time. An inability to synthesise evidence and listen to well-informed opinions from other experienced players is usually the hall mark of a weak player. If you’re playing CCGs and don’t understand stats, or worse: don’t believe them, you’re already at a huge disadvantage. If you value your own opinion higher than the aggregated opinion of the majority of the rest of the A:NR community then you have a clear disregard for the numbers.

If you had played 50 games of Netrunner every day since it came out, you’d have played something close to 23,000 games - which is still less than the sample we already have for Gabe on OCTGN. If your win rate with him was 10% better than the rest of the world you would boost the world average by less than 5% and Gabe would still be behind Andromeda. That’s how big a 6% edge is, and that also goes to highlight just how insignificant your “personal experience” is when it’s weighed against the integrated experiences of the rest of the world - a comically large edge and an unrealistic sample size still can’t save him, so your actual win rate and sample size would just be a drop in the ocean.

You already said your meta doesn’t travel, so you have no basis for comparison. How do you know that it isn’t just a case that your guys all play Corp really badly, which gives Gabe more of a comparitive advantage? How good really are your guys? How many regional or national placings have any of you achieved? Any pedigree from other CCGs that might add some credibility?

You can’t just write-off the OCTGN win rates. Yes, the online community might not be as strong as some local metas, but with sufficiently many games averaged over all ability levels that just doesn’t matter. If there was bias in the sample then it might - i.e. if all the Gabe players were systematically weaker than the Andromeda players.

Guess what? That might not be far from the truth! When players build a deck, they don’t do so in a vacuum - they don’t randomly choose an ID and then build a deck for it. They choose an ID based on what they believe to be successful and build their deck in the context of whatever else is popular in the community. The evidence, then, suggests that good players believe Andromeda to be a more successful identity than Gabe in the current popular environment. If your experience differs then perhaps your Corp meta is significantly different to the rest of the world too.

Even if we accept your hypothesis that Gabe is better at the elite level, there are still some very strong reasons why you would still choose Andromeda regardless. The runner already has an advantage over the Corp, it was pretty significant although the gap is closing. Many players might feel that they can do without the extra few percent that might be squeezed out of Gabe, and instead favour the consistency of Andromeda (she is lower variance, you cannot deny that). Moreover, that consistency makes Andromeda faster - which is very relevant at the KO stages of a tournament when agenda totals matter. I’d gladly sacrifice a couple of percent at the very top level (against players and decks I might not even get drawn against) in order to boost my chances in the far more common middle order games.

I would just like to point out that there definitely was at least 1 Gabe in the top 32 at Worlds. The BGG link to the top 32 decklists still doesn’t have every list. After this stimhack thread was made another player (Joe Held) added his lists to the thread. He was running Gabe.

Thanks for letting us know! Updating original post.

The environment does not need to be bad, just not the best players in the world. You are weighting your own groups anecdotal evidence really high. The problem is that obviously that is not really much in the way of evidence for those of us not in your group. :wink:

1 Like

I am very familiar with statistics, which is why I think issues of sampling and base rates are so important here. If you want to win an event, the average win rate of an identity across the broader Netrunner population shouldn’t be particularly important-- instead, you should be worried about how that identity does in high-level play.

A good example of this is Jinteki: Personal Evolution. This faction is quite strong against average to mediocre players, but is extremely weak against people who are good, barring unusual circumstances (like favorable elimination scores, as we saw at Nationals). Similarly, Weyland is very good at flatlining bad runners, but rapidly becomes less effective against skilled play. Looking only at Plugged-In, you might think “wow, Weyland is the best!” but this seems clearly wrong.

The “usual suspects” from the area have had very strong performances at regionals and Plugged-In events, as well as on OCTGN (though some of them no longer use OCTGN). I believe the standard of play here to be very high, and hope to attend Nationals or Worlds next year (schedule permitting) to prove it.

I think that there are several players in the local group who would have a good shot at winning Worlds if they attended.

I never get how that is even a discussion.
Gabe has an ability that telegraphs what he is going to do. Andromeda doesn’t.

That is what it comes down to for me. Knowledge is power, especially in netrunner and for the corp. Currently the two dominant identites are Andromeda and Kate. They don’t tell the enemy anything beyond the most basic stuff, Andromeda basically says “I am going to have a very consistent opening” and Kate says “I am going to install stuff”. That is no information given up.

On the other hand, most other identities give a whole lot more of information away just by being laid down on the table.
Kit: I am going to pressure you early with decoders.
Exile: I am going to recur some programs.
Gabe: I am going to run HQ.
etc. etc.

Of course this doesn’t describe everything that prevents an identity from being chosen by top players (noise lack of speed comes to mind), but to me that lack of flexibility is what makes some identities worse than the two currently at the top.

1 Like

The “high-level play” argument really doesn’t apply - in fact it almost never applies, unless you’re talking about an invitational tournament to which only the top players in the world are invited. The swiss stages of a tournament are not high-level play, they’re a mixed bag of whoever turns up on the day and, as we know, anything can happen. Good decks/players can lose to bad ones because the “bad” is employing sub-par but situationally useful cards which just happen to scupper the “good”; a freak draw leading to really swingy results; incorrectly assessing the meta etc.

Even if you happened to get drawn against an “elite” player in the first round, one of you has to lose (or draw) and so at least one of you would end up in the lower part of the field playing non-elites for a few rounds. As such, the “broader Netrunner population” is extremely relevant - these are the guys you will play day in day out, and they make their deck-building (and play) decisions based on popular opinion in the good majority of cases.

Also, if we’re to believe that Gabe has a higher win-rate at the highest level, the only explanation for the stats we’re seeing is that he has a significantly lower win-rate at the mid levels and that, I believe, is more significant than you give credit for.

It’s entirely possible that Gabe does have more potential at the highest levels (although I disagree), but I would still choose to play Andromeda. With Andy you gain significantly against “the field” - i.e. the average swiss match-up. The down-side (if there is one) is that you might give up a few percent against the very top players at the business end of the tournament. Andy reduces your chances of losing to a bad draw early on and improves your speed (which is important in the KO stages), whereas Gabe (if he has any benefit at all) only helps you in a small percentage of matches.

You’re absolutely right, but your argument doesn’t support this view. You’re saying that you believe your own experience more than the collective experience of the world. So your sample is narrow and limited, yet you put more stock in it than datasets much larger (and varied) than your own.

Take Worlds for instance, how many competitors were there? I don’t have the exact figures, but let’s say it was 100 for the sake of argument (since I know it was more than this). How many rounds of swiss were there, at least 5?

100 players for 5 rounds of swiss is 250 matches (500 games) + 31 matches (62 games) in the KO rounds. Have you played that much Netrunner over your entire lifetime? I’m willing to bet that the World Championships dataset on its own is comparable in size to your own personal experience data.

Finally, while on the subject of sample size, your “highest-level” claim is extremely hard to test. By definition this is a low statistics region of the sample. There are simply fewer elite players than there are mediocre ones, so the standard error for this portion of the player-space is much higher. How do you know that your personal observation of a high win-rate with Gabe isn’t just a random noise variation on top of a mediocre true win-rate evidenced by the larger samples elsewhere?

Research has been done by filtering OCTGN data by player skill (using a post-hoc ELO rating system or similar). Why don’t you check it out?

This is the crux of your argument: belief. I have no doubt that your local group is very strong, but you’re handwaving away all of the objective data here, man :). If you also believe in statistics, you can at most say that you have a doubt that Andromeda is as much better than Gabe as she appears, because of sampling problems with the data.

I cannot see a good faith/non-belief-based argument for Gabe actually being “better” in the sense that we’re talking about here: Andromeda is clearly superior at placing highly at and/or winning tournaments. Gabe’s special ability might well be stronger in some one-on-one matchups, but he’s simply not placing as highly in tournaments right now.

1 Like

Yeah, but if you’re good you don’t really need to worry about non-high level players, except in some special cases with matchup-dependent decks. You need to worry about the elites.

I agree with you-- I think Gabe is very easy to play badly.

I think it’s a bit short-sighted to say you don’t need to worry about non-elite players. Are you saying you beat average players 100% of the time? That would be quite a claim.

You’re always able to outplay inferior opposition and you’ve never suffered a bad draw? If you’re that good then you shouldn’t be worried about elite opposition, they would still only contribute a small blip on your flawless win rate.

No, but I very reliably “make the cut–” whether that means the actual cut or the de facto cut at the top tables of Swiss events. As such, I’m primarily concerned with what the good players are going to be throwing at me, not with improving my percentages against low- or mid-level opponents.

Are we positive on the 13/16 Andromeda? I was under the impression it was 12/16 (not that this would change much though)

Just looked at the BGG worlds thread, and it appears we had 11 Andromeda in the Top16 (4 Kate + 1 Whizzard)

People keep reporting different things.

Look like it was 11/16 Andromeda now. Out of those 11, 7 advanced to the top 8!

I find it odd that this discussion of Andy vs. Gabe is not discussing things like “what corp decks are you expecting to see”, “what specific corp cards are played highly/not played often in your meta”, and other questions regarding the relative strengths of the two factions at responding to specific corp strategies. I specifically switched from Andromeda to Gabe with the release of Second Thoughts, after I saw that a shift in the corp meta made Gabe, in my opinion, the better choice. As I’m starting to see a return to NBN and HB Rush decks, I may wind up going back to Andy.

Gabe and Andromeda decks are not built the same way, and thus it does make sense to think of them as doing the same thing, just to a different degree of effectiveness. For example, it is uncommon for Andromeda to play Sneakdoor, and when she does it is typically as a one-of. Gabe, on the other hand, usually plays 3 Sneakdoors. That sort of thing vastly changes how the two identities play, and thus what they are relatively strong against. Gabe is more able to cheaply land Emergency Shutdowns, for example. It also means he has less MU available for datasuckers: my current Gabe list does not run suckers, whereas if I were playing Andy at the moment there is no question that I would be running 3. Datasucker often does not provide enough of a discount to make brute-forcing a Hadrian’s Wall or Archer worthwhile, whereas against Bastion it is an allstar. By contrast, Sneakdooring past a Hadrian’s or Archer on HQ, then Shutting it down, is a huge game, whereas the corp can likely afford to re-rez a shutdowned Bastion. Similarly, Gabe’s ability pays off better over the course of a long game, making you more resilient to attrition strategies, whereas with only 5 cards in your opener, you’re less likely to see breakers/special orders, meaning a rush deck might be able to sneak an agenda or two past you.

I don’t know about the specifics of the corp-meta at Worlds, as a plane ticket to Minnesota is just not something to be found in a student’s budget, but I know the pre-Second Thoughts meta online was fairly hostile to Gabe; people were running enough ICE, and with cheap enough costs, that they would not feel overly strained defending 4 servers (whereas you typically just need 1 ICE on Archives to stop sucker-farming vs. Andy) and trying to sneak agendas quickly, specifically because they knew a late-game against the then-popular Atman was unwinnable. Given Andy’s success at Worlds, I am guessing a similar situation was present there.

2 Likes

[quote=“Arkhon”]I think it’s a bit short-sighted to say you don’t need to worry about non-elite players. Are you saying you beat average players 100% of the time? That would be quite a claim.

You’re always able to outplay inferior opposition and you’ve never suffered a bad draw? If you’re that good then you shouldn’t be worried about elite opposition, they would still only contribute a small blip on your flawless win rate.[/quote]

When we’re talking specifically about runner vs. corp games, I don’t think it’s a stretch to say I win 90%+ of my games vs. non-elite players. Because of how swiss pairings work, this usually means that I play the other top players at a tournament during the swiss, because they are also winning a bunch. I care more about big gains in my win percentage vs. high-skill players than small gains in my win percentage vs. average-skill players. This is not to say that switching from Andromeda to Gabe, or vice versa, would provide such gains, but I agree with Kingsley’s priorities.

1 Like

I really liked your post, but I think this part was off. In my experience Andy has a significantly stronger late game than Gabe. The RaD interfaces that she runs simply dwarf Gabe’s ability.

1 Like

Most Gabes run Indexing + Same Old Thing, though, which gives you just as much R&D inevitability, if not more.