The Rally Point (discussion and questions)

[stockwellpete]

To continue the subject of army allocation, I am now going to illustrate how a modified system might work using the FOG2DL Player Ratings and MikeC_81's army classification. See here for more details . . .

viewtopic.php?f=477&t=83417&hilit=MikeC_81
viewtopic.php?f=501&t=80847

I am going to use Classical Antiquity Division A from this season as an example to show the differences between the current system and the new one that I could introduce if players are generally happy with it.

The 10 players in the division are (with their FOG2DL rating in brackets and army choices after that) . . .
Nosy_Rat (no rating), Pontics, Romans, Jewish
TheGrayMouser (5.67), Gallics, Ptolemaics, Graeco-Bactrian
hidde (5.78), Carthaginians, Romans, Macedonians
NikiforosFokas (5.80), Thracians, Indo-Parthians and Ptolemaics
MikeC_81 (5.83), Romans, Carthaginians, Spanish
ianiow (6.00), Carthaginians, Macedonians, Ptolemaics
shadowblack (7.09), Spartans, Romans, Macedonians
rbodleyscott (7.33), Rhoxolani, Romans, Carthaginians
ruskicanuk (10.00), Carthaginians, Romans, Carthaginians (Achaemenid Persians)
klayeckles (10.33), saka, Indo-Skythians, Ptolemaics

Now under the current system I would look at that pattern of army choices and try to maximise the number of first choices without any regard to the player ratings, or even the player's names. And with this group I was able to come up with 8 first choices (:D) and one second choice. One player who squeezed into the division right at the end of the recruitment process was asked to make another third choice because they had chosen two armies from the same nation in their original selection.

But using the alternative idea, where the "order of preference" is removed, I now have to look at the FOG2DL ratings and allocate armies to the lowest rated players first. So the first player is Nosy_Rat who is an unrated player. This means he is either new to the FOG2DL, or has not played the requisite number of matches (15) to get a FOG2DL rating. Normally, I would not put a player straight into Division A but his form in other tournaments suggested he would probably cope OK. His choices are Pontics (which are a bit of a non-descript army), Romans and Jewish. I am not sure where the Pontics come in Mike's classification but roughly I would rate them as tier B/C so, given that Nosy_Rat is bit of an unknown quantity I think I would give him the Romans, because they are a better army than the Pontics.

Next up is TGM who would be given the Gallics and then comes hidde who would get the Hannibal/Carthaginian army. NikiforosFokas would get the Thracians but MikeC_81 would have to be allocated the Spanish, as the Romans and Carthaginians have already been taken. ianiow would get the Macedonians because the Carthaginians have been taken, shadowblack would get the Spartans and rbodleyscott would have to have the Rhoxolani as the Romans and Carthaginians are already taken. ruskicanuk made only two valid choices so was asked to make another selection and chose the Achaemenid Persians and klayeckles would be given the Ptolemaics in preference to the Saka because the Ptolemaics are a better match-up against the other armies in the division. For all of these divisions I did not need to consult Mike's guide as I am dealing with experienced tournament players who know their armies very well.

So the final line-up using the alternative system would look like this. Seven players ended up with the same army while Nosy_Rat, MikeC_81 and klayeckles got something different (actual army used shown in brackets) . . .
Nosy_Rat - Romans (Pontics)
TheGrayMouser - Gallics
hidde - Carthaginians
NikiforosFokas - Thracians
MikeC_81 - Spanish (Romans)
ianiow - Macedonians
shadowblack - Spartans
rbodleyscott - Rhoxolani
ruskicanuk - Achaemenid Persians
klayeckles - Ptolemaics (Saka)

Because this method allows me to be a bit more interventionist, I was able to give a little bit of a help to the lowest-rated (non-rated) player in the division thereby reducing the chances of having a player tailing off after a few matches if the standard was a bit too tough for him (it wasn't as it turns out ) and I was able to exclude a horse archer army that didn't really fit in with the other armies (it is arguable, I know).

I was hoping to find out from this exercise whether players would need to choose 4 armies instead of three, But it is still a bit inconclusive. I will have to do another test later on today or tomorrow. My guess is that we will need to ask for 4 choices with this method to cover all eventualities.

[stockwellpete]

Ah you mean it wasn't programmed to only allow rear charges if the unit's whole turn started behind the flank/rear?

Yes, I think that was it.

[stockwellpete]

Yes, I didn't know about it as an option. I have probably missed it in the rules.

I am bringing it in for Season 3.

[MikeC_81]

In terms of anomalies, so far I have picked up one for the hybrid 2/3 idea and four for option 4 in the poll. They are . . .

1) hybrid 2/3 idea - players in a tactical 35% -35% draw will score 1 point each, the same as a player losing 20-44 in a marginal defeat.

2) option 4 - a player losing a match 20-45 will score 1 point, the same as two players drawing a tactical match 39-39.

3) option 4 - a player losing a match 40-65 will score 2 points, more than players drawing a tactical match 39-39, and more than a player drawing a match 39-15 or 39-40.

4) option 4 - a player losing a match 40-65 will score 2 points, the same as players in a 59-59 draw.

5) option 4 - a player losing a match 0-40 will score 0 points, the same as a player drawing a match 19-0.

I personally think we should be less concerned with anomalies and more concerned with eliminating any chances that a player is given incentives to avoid combat. I would even be in favour of giving losers a point and draws 0 points just to encourage getting results.

At the end of the day, I couldn't care less about whatever the minimal prize support there is out there and whether any potential prize I could win is given to someone else because of some scoring anomaly. I am playing, and I hope the majority of those playing, are doing it because it is fun and we get to explore the full potential of the armies in the game pursuing the built-in victory conditions of the game. Playing to try and win while someone else is merely trying to deny you a win, instead of themselves getting a win, is probably the most unfun experience possible.

[TheGrayMouser]

Ah you mean it wasn't programmed to only allow rear charges if the unit's whole turn started behind the flank/rear?

Yes, I think that was it.

Yes it was a bug introduced by a patch I believe was to fix some minor line of sight issue. It was insidious because it only worked if you moved a unit hex by hex. I suppose the players that used it felt justified because of the “ if the program allows it it it’s ok” mentality, even if their opponents were clueless. Put on damper on my enjoyment of competition play.

[stockwellpete]

I personally think we should be less concerned with anomalies and more concerned with eliminating any chances that a player is given incentives to avoid combat. I would even be in favour of giving losers a point and draws 0 points just to encourage getting results.

At the end of the day, I couldn't care less about whatever the minimal prize support there is out there and whether any potential prize I could win is given to someone else because of some scoring anomaly. I am playing, and I hope the majority of those playing, are doing it because it is fun and we get to explore the full potential of the armies in the game pursuing the built-in victory conditions of the game. Playing to try and win while someone else is merely trying to deny you a win, instead of themselves getting a win, is probably the most unfun experience possible.

Vote for option 2 then. Along with the new 2-turn re-start regulation we should be making a big impact on negative play in Season 3. You will have to inflict 20% damage on an opponent to start scoring points in future.

[rbodleyscott]

and rbodleyscott would have to have the Rhoxolani as the Romans and Carthaginians are already taken.

Noooooooo! Not again!

[rbodleyscott]

I am pretty sure that I don't like the idea of the tournament organiser picking what he deems to be to the best army from their selection for anyone. Leaving aside the issue of subjectivity, perhaps a player wants to try out a more challenging (dare I say more interesting) "off-meta" army but has to fill out his choices with sensible armies because he does not want to get stuck with a challenging army that he does not think he will enjoy playing. And then the tournament organiser selectively picks one of the the "sensible" armies for him, so he might as well not have bothered to include the "off-meta" one in his choices. This could be particularly annoying if the player reckons he has discovered how to win with an "off-meta" army, but is denied the chance to try.

Would the unexpectedly successful Thracians and Scots-Irish have even seen the light of day last season if the proposed system had been in force? (Following the success of the latter, nearly half the players in Division A opted for swarm armies in the Late Antiquity section of season 2. So from "off meta" they went to "on meta" which might never have happened under the proposed system. I suspect they may be "off meta" again by next season).

It isn't always about minimaxing. Do you think I thought my choices for the current league were the armies with the best chance? Not everyone wants to take the top-rated army. They may just want to try something a bit different and see how well they can do with it. In the Rhoxolani case this turned out to be hubris, and it looks like I will be relegated to Division B, but you live and learn, and enjoying the games and trying something different is more important to many (most ?) players than promotion or relegation.

The only fair way to do it without order of preference is completely randomly.

Either way, it seems that for the highest ranking player, it is a bit like the TV show "Pointless". You are trying to find the best army that nobody else has picked. Because if anyone else has, then you are not going to get to play with it, particularly if the tournament organiser is selectively allocating it on the lower-ranked players' behalf, even though they may have only included it in their choices because they had to fill out their choices with something.

Ideally tournament systems should not involve any judgement calls by the organiser whatsoever. If this cannot for some reason be achieved, then such judgement calls should be minimised, not given a major role in the system.

[Kabill]

To continue the subject of army allocation, I am now going to illustrate how a modified system might work using the FOG2DL Player Ratings and MikeC_81's army classification. See here for more details . . .

I don't think I'm keen on this system. There's two key reasons:
1) Players may not prefer to play the strongest army of all those they have selected. I'm quite possibly in a minority here, but I would order my picks based on what I feel like playing, not what is the "best" of the choices I have made, and so under this system if I wanted a shot at playing the armies I really wanted to play, I'd have to avoid including any stronger picks (even if in principle I would be happy to play them if my others were not available) as those would reduce my chances. So I'm more likely to be dissatisfied and also have a incentive to "game the system" to get one of the armies I particularly want.
2) I think there will be too many instances where you're having to apply personal discretion to decide which army a player will use. E.g. What do you do with lists that there isn't a rating (e.g. the Pontics in your example), which would be a major issue for sections based around new DLC periods? To what extent will judgements about army ratings take into account rules changes/patches (e.g. I don't know if MikeC's list has been updated considering changes to how massed bowmen work with the recent patch but if it hasn't, are you going to take that into account)? Etc. To be clear, I am personally convinced you would make decisions in good faith with a view to making the tournament as enjoyable as it can be for all participants. But given that there's at least some degree of subjectivity in the ranking of army lists, and inevitably a number of cases where you'll have to apply discretion, I'm concerned this would afford more scope for players being unhappy because they disagree with your decisions, and more scope for people accusing you of bias or preferential treatment.

In contrast, the system you have at the moment involves minimal use of discretion and maximises the number of players who get to play with armies they particularly want to play with. For the sake of trying to deal with the relatively minor problem of draw games, I don't think it's worth throwing away what seems to be a pretty good way of organising things.

EDIT: Ninja'd by RBS

[stockwellpete]

I am pretty sure that I don't like the idea of the tournament organiser picking what he deems to be to the best army from their selection for anyone. Leaving aside the issue of subjectivity, perhaps a player wants to try out a more challenging (dare I say more interesting) "off-meta" army but has to fill out his choices with sensible armies because he does not want to get stuck with a challenging army that he does not think he will enjoy playing. And then the tournament organiser selectively picks one of the the "sensible" armies for him, so he might as well not have bothered to include the "off-meta" one in his choices. This could be particularly annoying if the player reckons he has discovered how to win with an "off-meta" army, but is denied the chance to try.

I think you are over-exaggerating the input I would have in the army selection procedure I am suggesting. Looking at the example that I gave of the new idea in operation, I ended up picking 2 of the 10 armies. And these were picked from a small group of armies that the players had already indicated that they wanted to use. So the scope for "subjectivity" is extremely limited. One was for an unrated player where my concern was that I might have put him in a division that was too high for him, the other was the last army choice where I felt that the Ptolemaics would give far better match-ups overall than the Saka. Both decisions were made for the benefit of the tournament overall, particularly in light of current concerns about negative play. A player outclassed by the others in his division is more likely to "turtle" his army than other players, and horse archer armies are disproportionately involved in sterile draws. Everything else was chosen from the first army the player had put in their list even though there was no formal "order of preference". Perhaps, in future, this system would develop a "nod and a wink" order of preference, rather than the current formal one, where the first-named army in the list is understood to be the one the player prefers the most.

Would the stunningly successful Thracians and Scots-Irish have even seen the light of day last season if the proposed system had been in force?

Yes, as the Thracians were in the only example of the method at work that I have posted so far.

It isn't always about minimaxing. Do you think I thought my choices for the current league were the armies with the best chance? Not everyone wants to take the top-rated army. (In the Rhoxolani case this turned out to be hubris, but you live and learn).

I don't see how this point is relevant at all. I imagine players choose armies for all sorts of reasons, all equally valid in my opinion.

The only fair way to do it without order of preference is completely randomly.

Interesting. Please explain. Do you mean something like this? I get my old trusty bingo kit out and put the balls numbered 1 to 10 in a bag and pull the first one out. That will give the player's name. Then, in another bag I put another 3 or 4 balls and pull another ball out and that gives me the army for that player. Then I move onto the next player and so on. I think this might work and the only question again is how many army choices players initially would have to make. Almost certainly 3 would not be enough, maybe 4 would be? That would be completely impartial and would be within the range of ideas that I could support as the organiser. It would remove any discretion I could exercise though for the benefit of the tournament. In the FOG2DL I have not entered sections where I have to allocate armies and I have now retired from competitive play so I am a completely neutral benevolent despot.

Either way, it seems that for the highest ranking player, it is a bit like the TV show "Pointless". You are trying to find the best army that nobody else has picked. Because if anyone else has, then you are not going to get to play with it, particularly if the tournament organiser is selectively allocating it on the lower-ranked players' behalf.

But who is the highest ranking player? It was klayeckles at the start of the season but it's not him now. There is plenty of variety in the FOG2DL so players generally get quite a bit of what they want. Otherwise, why do they keep coming back?

Ideally tournament systems should not involve any judgement calls by the organiser whatsoever. If this cannot for some reason be achieved, then such judgement calls should be minimised, not given a major role in the system.

Yes, I basically agree. What you are being asked to weigh up here is whether a very modest amount of intervention by me in the area of army selection is worth it if it helps to reduce negative play and increases the number of interesting match-ups in the tournament.

Just to be clear, I am more than happy with the current system, but it seemed encumbent on me to suggest alternatives given there is quite a bit of concern at the moment about negative play. So far we seem to have three ideas. The existing system, the benevolent despot system and the bingo system. I can feel another poll coming on.

[stockwellpete]

I don't think I'm keen on this system. There's two key reasons:
1) Players may not prefer to play the strongest army of all those they have selected. I'm quite possibly in a minority here, but I would order my picks based on what I feel like playing, not what is the "best" of the choices I have made, and so under this system if I wanted a shot at playing the armies I really wanted to play, I'd have to avoid including any stronger picks (even if in principle I would be happy to play them if my others were not available) as those would reduce my chances. So I'm more likely to be dissatisfied and also have a incentive to "game the system" to get one of the armies I particularly want.

2) I think there will be too many instances where you're having to apply personal discretion to decide which army a player will use. E.g. What do you do with lists that there isn't a rating (e.g. the Pontics in your example), which would be a major issue for sections based around new DLC periods? To what extent will judgements about army ratings take into account rules changes/patches (e.g. I don't know if MikeC's list has been updated considering changes to how massed bowmen work with the recent patch but if it hasn't, are you going to take that into account)? Etc. To be clear, I am personally convinced you would make decisions in good faith with a view to making the tournament as enjoyable as it can be for all participants. But given that there's at least some degree of subjectivity in the ranking of army lists, and inevitably a number of cases where you'll have to apply discretion, I'm concerned this would afford more scope for players being unhappy because they disagree with your decisions, and more scope for people accusing you of bias or preferential treatment.

Please see my answers to Richard's post as they are covering much of the same ground. If we take that hypothetical Classical Antiquity A Division example again, the only player that I intervened on behalf of was Nosy_Rat because he was an unrated player. Although the ratings of the other players ranged from 5.67 to 10.33 I think they are all capable of beating each other on their day, so there was no need for any "subjective" input by me. I have said that I would only be using MikeC_81's army classification as a guide, and all that means is that I wouldn't be giving a player like Nosy_Rat in that example a tier C army. I would try and give them the strongest army they had chosen to give them a fighting chance. Now that I have seen Nosy_Rat in action for a season I can say that there will be no need for me to intervene in a similar fashion on his behalf in future. For the other 9 players in the division I didn't even consider Mike's classification. There was no need to as I know all of them to be experienced players who know their armies. The point of the new system is largely to do with giving me some extra flexibility to deal with awkward situations that occasionally do occur. Sometimes I will look at a division and think, "so and so is really going to struggle there" and it will enable me to do something about it. The alternative is the increased likelihood of negative play (it doesn't always happen but it might) or even a player dropping out.

In contrast, the system you have at the moment involves minimal use of discretion and maximises the number of players who get to play with armies they particularly want to play with. For the sake of trying to deal with the relatively minor problem of draw games, I don't think it's worth throwing away what seems to be a pretty good way of organising things.

Yes, that's the issue. The current system is perfectly OK. It just seems sensible to raise alternatives every now and then in order to keep the tournament up to date and fresh.

[MikeC_81]

Just to be clear, my tier list is strictly a guide based off *known* results from competitive play and where required my own personal experience and analysis when an army has not seen a lot of play.

It is more of a guide for players to avoid getting ambushed by absurd match ups in pick up games or as a guideline for a newer player trying to enter the FoG2DL or other player run tournaments but do not have enough experience to spot trap armies which are bad.

It should not be used to allocate armies because different armies require different skillsets. If you hand a newbie the Carthage armies, they are likely to get crushed because there is a a high skill floor before that army becomes competitive. Meanwhile, a player can literally blunder into a win off the backs of Veteran Legionaires in the Roman lists

[rbodleyscott]

Forcing everyone to play with competitive armies "for the greater good" is not the answer to negative play. Well it is an answer, but one that will cause a lot of resentment, and probably reduce the overall variety in the tournament.

In any case, it would not solve the problem, because several highly rated players (whose choices you have said you would not manipulate) have entered with dodgy armies in one or other section in season 1 and/or 2, had their asses handed to them in the first few games, then played negatively in later games.

Even with a "competitive" army, a player who is intimidated by his opponent's reputation is just as likely to turtle up. The only difference being that he is more likely to do so successfully, whereas with a dodgier army his opponent might have felt able to go for it despite the enemy advantage of position.

[pantherboy]

Just got back from Okinawa so forgive me for being late to the discussion. In my opinion the current scoring system is open to gamey play. Across my 27 matches this season I would say a half or just over half were players sitting in the best defensive position looking to earn a draw or expecting me to walk into obvious traps. This was irrespective of whether they had the better army or not. Some even stated on turn one that they wanted a draw and would be happy with that. The only match I could not force was versus Ironclad as the battlefield made it impossible for either player to attack without being slaughtered. The matches I lost or I had a tight run were only with players who actively maneuvered. Imagine going to a soccer match were one team is rated better than the other so the weaker team forms an 11-player tortoise with the ball in the center and then they do not move so as to score a draw or push it to a penalty shoot out. If you do not fight a battle then you should not earn any points at all as no battle occurred for a draw to result. Once again imagine two chess players simply sitting down and doing nothing more than agreeing to a draw without moving a piece. You would not call that a game. And to talk about losses coming from pursuit or clever positioning for a draw or whatever to justify such play or scoring systems is rubbish in my opinion. These are meant to be competitive matches to rank players upon performance. The goal is to do your best with the army at hand whether you selected it or not. That is why I used the following scoring system for the Pike and Shot tournament that I ran and which Slitherine have picked up for their tournaments. The problem with the way Slitherine are using it is that the match samples are too small and you get blowouts from mismatches. A skilled player could face a relative novice in the third round by simple fortune. In a group where you play everyone then the system shines. I feel winning a 60% to 59% match where the simple cast of the die may have decided the winner should not be an all versus nothing score. I feel a player who looses every match but pushed every one of their opponents to within a percentage point or two of breaking must of played well and deserves to have that reflected in their tournament scoring as compared to someone who wins one match but looses everything else dismally who would be rated higher. The idea of playing for a draw to stop being demoted is insane. If you do not deserve to be in a lower bracket than the next season you will trounce your opponents and be promoted back up. This scoring system encourages you to fight a battle regardless of the conditions or match up as the better you do the more points you will earn. I am not implying that you can not take up strong defensive positions but doing so with the hope of dissuading your opponent to attack and concede a draw should not be the motive. In such a case you should both restart the match as many times as necessary till one of you feels that they can force a result. The goal is to have a competitive match pitting skill versus skill (plus luck) so that you can get a fair approximation of where you stand in the rankings.

5. Scoring

Scoring will be based upon the break level and by how close the battle was. The narrower the margin between players will result in scores rather close while an early decisive win will be wider.

Win: 6.0 + (enemy break% - winner break %)/10
Draw or Loss: (enemy break %)/10

Example:
pantherboy (German Catholic 1632-1638) (61%) vs flatsix518 (Swedish 1630-1634) (48%)
pantherboy (48%)/10 = 4.8pts
flatsix518 6.0 + (61% - 48%)/10 = 7.3pts

NB: Any battle that ends due to all the turns elapsing and with a player ahead in break % will not be considered a win and as such will only score the opponents break level divided by 10.

[stockwellpete]

Forcing everyone to play with competitive armies "for the greater good" is not the answer to negative play. Well it is an answer, but one that will cause a lot of resentment, and probably reduce the overall variety in the tournament.

In any case, it would not solve the problem, because several of the most highly competent players have entered with dodgy armies in one or other section in season 1 and/or 2, had their asses handed to them in the first few games, then played negatively in later games.

Even with a "competitive" army, a player who is intimidated by his opponent's reputation is just as likely to turtle up. The only difference being that he is more likely to do so successfully, whereas with a dodgier army his opponent might have felt able to go for it despite the enemy advantage of position.

I don't understand why you are saying this. Players make their own army choices. I am not forcing anybody to do anything. It is nothing to do with me until I have to form up the divisions.

[stockwellpete]

Just to be clear, my tier list is strictly a guide based off *known* results from competitive play and where required my own personal experience and analysis when an army has not seen a lot of play.

It is more of a guide for players to avoid getting ambushed by absurd match ups in pick up games or as a guideline for a newer player trying to enter the FoG2DL or other player run tournaments but do not have enough experience to spot trap armies which are bad.

It should not be used to allocate armies because different armies require different skillsets. If you hand a newbie the Carthage armies, they are likely to get crushed because there is a a high skill floor before that army becomes competitive. Meanwhile, a player can literally blunder into a win off the backs of Veteran Legionaires in the Roman lists

Nevertheless, I still feel able to use it as a guide. In my example, the Romans are clearly a better bet than the Pontics for a player who is new to the FOG2DL and is being put straight into Division A.

[rbodleyscott]

Forcing everyone to play with competitive armies "for the greater good" is not the answer to negative play. Well it is an answer, but one that will cause a lot of resentment, and probably reduce the overall variety in the tournament.

In any case, it would not solve the problem, because several of the most highly competent players have entered with dodgy armies in one or other section in season 1 and/or 2, had their asses handed to them in the first few games, then played negatively in later games.

Even with a "competitive" army, a player who is intimidated by his opponent's reputation is just as likely to turtle up. The only difference being that he is more likely to do so successfully, whereas with a dodgier army his opponent might have felt able to go for it despite the enemy advantage of position.

I don't understand why you are saying this. Players make their own army choices. I am not forcing anybody to do anything. It is nothing to do with me until I have to form up the divisions.

You have said that you will select the "best" of their choices for "some" players. You are effectively forcing those players, however few they may be, to play with the army you deem "best" (out of their choices).

If you are not going to do that, then I don't know what we are discussing, because that seems to be the heart of your proposal.

[pantherboy]

In regards to army allocation when I ran LOEG I avoided army duplication in divisions so as to add greater variety to match-ups. As such I simply asked players to submit three armies by preference and awarded them on a basis of past performance. The lowest rated player received their first pick and so on up the chain. This invariably led to most first preferences except for the players at the top of their division who may of ended up with their second or in rare instances third preference. I agree that the umpire shouldn't attempt to deduce anyone's wishes. If a group ends up with some powerhouse list for a player than so be it. Let the chips fall as they may.

[rbodleyscott]

That is why I used the following scoring system for the Pike and Shot tournament that I ran and which Slitherine have picked up for their tournaments. The problem with the way Slitherine are using it is that the match samples are too small and you get blowouts from mismatches.

Agreed, but the tournaments are not aimed at a hard core audience who want to play against 9 (or whatever) different opponents. Equally, not being hard core, they are less concerned that only the best of the best have a crack at placing. In fact, the fact that a lucky draw could give more players a chance is actually a bonus for a more casual audience.

Anyway, there is nothing wrong with your scoring system per-se, it remains a major advance.

The main issue I think Pete had with it was the BYE score that I allocated, but that would not really be an issue the way that the FOG2DL is organised, with much more time to complete games, and Pete actively policing and replacing dropouts.

[Ludendorf]

I would stick with just letting players pick their armies by order of preference. Maybe check in with newer players to alert them to the fact their army has a bit of a bad record by the statistics, but that's as far as I'd go, and even that I'm leery about because it may unnerve new but talented players out of picking an army that nobody else has figured out how to use well, but which they actually have a really good strategy with. I'm quite comfortable playing with a 'low-tier' army if that's what I want to do; generally, if a player wants to take that kind of army into the league, they have some experience using it to compensate, or they just want to challenge themselves.

If someone wants to take an army out onto the field that's statistically a bit sub-par, I say let them. They may well give the other players a bit of a surprise in the event!