The section that I emphasised in bold contradicts that which you wrote in the reply which precedes this one. Again, therefore, I think that it would be beneficial for you to clarify whether you still stand by the statement in bold:
Theoryhammer in the sense you're describing is bad, since it's almost always results oriented thinking, i.e. players forming opinions of units by trying them out without actually doing any calculations to see what they're actually good at beforehand. Up until the last few years, the overwhelming majority of "tactic" or "strategy" articles on 40k were almost purely anecdotal nonsense, and the only calculations in them ignored points values entirely. Ask yourself many times you've seen a discussion get bogged down in "they're T3, so they're bad" or "they have a 3+ save, so they're good", as if that was relevant without considering the whole unit in the context of its points cost.
I have no idea why you think there's a contradiction. The bold text is how I described results-oriented thinking. That results-oriented thinking generates conclusions from direct experience with limited data sets (i.e., your own games), and sharing those conclusions generates anecdotes. Those anecdotes don't have good predictive value because they only show what happened, not what could happen and how likely the various outcomes are, so they aren't nearly as useful for informing future decisions as probability calculations. Hence, I labeled articles using results-oriented thinking, and especially in the absence of calculations based on points efficiency, as "anecdotal nonsense". None of what I said is incompatible with the idea that our learning is enhanced by sharing information, which is why I agreed with you on the importance of doing so.
In addition, your assertion that anecdotal experience is nonsense is, at best, open to question. I've learnt far more from reading the advice posted by players who have posted their experience of fielding units on the table, discussing the results, and working out whether problematic units could be used more effectively or are not overly worthwhile including in the majority of army lists.
...
Gaming experience wins over maths every time. I've read many a discussion about the maths underpinning unit performance and they have never told me anything that I hadn't already worked out from playing the game.
...
Maths has its place, but all it reliably provides is an indicator of performance of a unit in isolation.
As I said before, you took "anecdote" to mean something much broader than I do; you seem to mean it as "[almost?] everything except maths". You then write about the importance of gathering data and sharing it. Fine, but I think that approach is insufficient on its own. You then run into some contradiction of your own in the words I've highlighted in red because what you're describing is mathematical comparison. Even if you don't explicitly mention any numbers in that process, the comparison you're describing is rooted in judgments of relative value in a subset of specific game states with limited variables ("unit in isolation", but I'll forgive your hyperbole because you clearly don't mean it). When you draw conclusions about whether a unit would be useful in a majority of lists, you're doing it based on the perception of probability that a given cluster of variables in the aforementioned game states will actually occur in future games. That's exactly the kind of modelling you said had little value, and you steadfastly resist kind of quantification that makes
what you're already doing yield more precise results.
I'm not making any demands. I am demonstrating to you that model making is not a be all and end all of playing this game and that more often than not it is fundamentally misleading because of all the omitted variables. In addition, I argue that all this unnecessarily complicated mathematics is not required to understand how to play this game well. Basic probability is useful and working out various averages can be handy too, but ultimately it's in game experience of how to use units effectively in conjunction with each other which, in my experience of gaming and reading, is the key to playing all armies well,
As I just explained above, you and I are advocating the some approach except 1) that your aversion to writing out the numbers had led to your rejection of their usefulness, and 2) that rejection prevents you from quantifying a given unit's expected performance, so you have to rely solely on in-game data from your own experience or that of others. Sometimes I worry about the degree to which you and I get stuck in these debates as if our positions are somehow mutually exclusive.
Which is why I would never consider actually doing it. This level of modelling simply isn't required for a game. In addition, the whole point of 40K Online and other forums is to share experiences and ideas, not to carry out in depth research and modelling. Of course, if anyone really wants to engage in such research they can, but it tends to have a limited audience, owing to the complexity of the maths involved.
I was disputing the claims that 1) there were game elements that could not be modeled, and 2) that showing that all elements could be modeled required actually building comprehensive model.
I'm not sure where you get the idea that methods of probability or other mathematical modeling/simulation are somehow less legitimate/appropriate pieces of information to share on forums. The internet is packed with people asking questions like "How can I do X better [or at all]?", and many of the answers to such questions involve presenting new tools and methods that make it easier to solve both current and future problems. Sure, it takes some effort to learn how to use those tools and methods, but the payoff is worth it. The pool of human knowledge is as deep as it is because we share our tools and methods; instead of re-inventing the wheel, we get to benefit from the collective effort of millions of individuals, each of whom might contribute an idea that affects the whole population.
For example, I recently learned about the Lanchester square laws. Knowing them years ago would have saved me a huge amount of time. Similarly, I've been working an a spreadsheet to serve as a kind of developing tool for tinkering with 40k unit loadouts so I can provide better feedback to GW, or possibly design my own game at some point. I had no idea how to write 80% of the formulas in that sheet less than a month ago, so I looked at a similar file made by Neil Philips shortly after 8th edition was released. I wanted to write a spreadsheet anyway, but it would have taken me a lot more time to learn how to do it if I didn't have Neil's work as a starting point.
Regarding the small audience, after being away from TheWarmaster40k for almost two years, the first post I saw in their Eldar forum upon my return was by a player who admired my "The Mattler's Mathhammer" series and had written their own article in a similar style, listed me as a source of inspiration, and even quoted me in one of the preamble sections. I've been doing what I do for a while, and I'd still do it without an audience, but I learn a lot in the process of sharing it with others. It also seems to be making a difference to at least one other person, and that's a nice bonus.
Getting back to the issue of Dire Avengers. I think that it's clear that few would argue that they are not currently overpriced, however, I think that it remains very much open to question how much they should actually cost. I've seen nothing here that persuades me that they should be 13 points per model. I think that we need a lot more battle reports involving Dire Avengers to be posted to have a better idea of working out their cost. This is how we used to operate for the Eldar Redux, that is to say we played a lot of games to come up with a points value, so if anyone wants to make a convincing case for Dire Avengers to be costed thus, I'd recommend actually playing a number of games using this value for Dire Avengers and then seeing whether it's justified. That, to my mind, makes for a more more compelling argument and constructive discussion.
Fair enough. Even tools like the Lanchester square formula need their results to be viewed in the context of factors it doesn't (always) take into account, especially movement, and your opponent's might surprise you with a situation you haven't considered. Something to keep in mind, though, is that it isn't the Dire Avengers you'd need to keep an eye on during play testing. Their performance per model on the table will be the same regardless of their points cost. The real benefit to having a lower points cost is being able to field more units, which are usually different from the cheaper unit being play tested. An extra 40pts per full Dire Avenger squad can add up quickly if it means extra models or better armament for something else.
In addition to this earlier answer, it seems to me that armies have become more expensive in eighth edition. There are fewer units on the table than in other recent editions and when factoring in relative cost across the Eldar army and other armies, a reduction to a points cost based on editions employing far more complex rules seems, to me at least, difficult to justify.
That's a good point, which is why I based my argument for a lower points cost for the Dire Avengers on the current cost of a Tactical Space Marine because 1) it's the standard infantry model of another elite faction, 2) its points cost remained stable from 7th to 8th despite other units becoming more expensive, and 3) both models fill very similar roles in their respective factions. Seems like an reasonable comparison to me.
@Irisado- I bow as ever to your debating skills Irisado.
Heh, laying it on a bit thick there. However, I will say that Irisado maintains the highest standard of professionalism of anyone discussing 40k on the internet. Seriously!
I totally agree with you on math-hammer failing to tell the whole story.
It can, but it almost never does. You can make it super complicated if you want to account for every contingency, but for forum posting or even article writing it's not necessary. (...says the guy building universal hit/wound/save/damage formulas in Excel, including all special rules.
)
Anything less than using their own benchmarks to prove a point sounds more like whining than valid argument, and no one is going to listen.
Well, the points costs themselves are as much their own benchmarks as the power system, and a ton of their power costs are way off as well, and not just relative to points costs. The Wraithknight sword build is a huge ripoff because the power cost is scaled for the heavy guns, and both power and points costs of everything on the model probably need re-evaluation in the context of Imperial Knights. They're close, but not close enough to avoid suspicion. If you want to see a really stark contrast on power for points costs (other than Dire Avengers, who are great for power but terrible for points), compare Warp Spiders to Shadow Spectres.