Has anyone ever seen an explanation from the designer of what design problem he felt was solved by the AoS income reduction mechanism? If (as would seem obvious) the goal was to alleviate some of the "rich get richer" issues with the game and give the trailing players a slim chance of catching up, then why is the AoS version of the rule so much less effective at accomplishing that than the one that appeared in Lancashire Railways? This version not only is bad at helping the trailers (most often, they benefit rarely), but it also introduces extremely gamey behavior such as "helping" a fellow player by riding on them to push them over a boundary on the income track.

I'm just curious what went into the design.

Randy...

I haven't seen one.



That isn't obvious to me at all. To me the income reduction mechanism is a rather startlingly elegant system for ensuring that players have to continue to work hard in order to progress; that the hill remains steep and the game does not necessarily become inherently easier the more successful you are. As all players have to climb this increasingly steep trail, and will hit the same steep slope, usually harder, when they get to that same point makes the claims of hindering the leader seem questionable at best.



Yeah, the 10 dollar cliffs are a bit odd. I've been playing lately with removing it in some of my designs (income reduction ever $5, starting at $6) and it really has very little effect other than delaying the end-game by about a turn (see recently posted images for AoS:Wales and AoS:Launch!). I imagine that Martin added the $10 cliffs simply because they added another interesting decision/risk for players to manage in the game. Certainly that would have likely been my reason in his place.

Martin's design submission "The Winsome Rail Game" that became "Lancashire Railways" had no income reduction. The income reduction mechanism in Lancashire Railways (see "Determine and Record Market Changes") was created in 1998 during development here in Pittsburgh. Regarding the movement to the Income Reduction mechanism developed for Age of Steam, the reasons are:
>Simpler, Faster, easier for the gamers. Brain dead simple, yet effective.
>Fixed, Predictable, less randomness. We were charged with developing a 'gamer's game'.
>Interactive, Player Influenced. Yet another capability for player 'screwage' added to the overall player interaction.

Have fun!

John
Pittsburgh

and in case anyone is trying to follow along at home:

The Early Railways games also used an income reduction system, but it was dictated by a roll of the dice instead of being at a certain, set level.

Respectfully, effective at what? Was the goal, in fact, to slow down leaders so that early misfortune or mistakes could potentially be recovered from?




This sounds like another reason beyond reining in runaway leaders. I'm uncomfortable with "interactive" mechanisms that are disproportionately harsh on trailing players, but I can see how tastes might differ.

Randy...

It is not even a runaway leader that is the problem, but the fact the game has extreme runaway exponential growth. Because each shipment increases your income permanently, as the game continues, your abililty to increase your income improves (longer link count, more routes, etc.) Income reduction steepens the hill meaning you have to keep increasing THE INCREASE in your income for your income (& score) to keep going up.

I apologize for the extreme highlighting--but I think the distinction is extremely important.

This table shows what happens if Income Reduction were not there. 8 turn game, player always upgrades their train on the first shipment (Locomotive is never taken) and make 2 shipments once the train reaches the maximum 6 link level and is always able to ship a cube their maximum links every time:

| Income reduction | No income reduction
Turn L | SI PSI FI G | SI PSI/FI G
1 2 | 0 2 2 2 | 0 2 2
2 3 | 2 5 5 3 | 2 5 3
3 4 | 5 9 9 4 | 5 9 4
4 5 | 9 14 12 3 | 9 14 5
5 6 | 12 18 16 4 | 14 20 6
6 6 | 16 28 24 8 | 20 32 12
7 6 | 24 36 30 6 | 32 44 12
8 6 | 30 42 34 4 | 44 56 12
L = Links, SI = Starting income for turn, PSI = Post shipment income,
FI = Final income for turn, G = Income level gain

Under income reduction, except the big bump when a player starts to deliver two 6-link cubes per turn, the gain in their income is relatively flat--and it drops quickly to its pre=bump level.

Under no income reduction, the game literally explodes out of control.

BTW, even if you were allowed to upgrade past 6 under this scenario, the player would not do it, unless there were no other cubes of length of 5 or 6 to deliver. The time (and increase in income) lost in the upgrade does not get replenished before the end of the game.

Income Reduction is NOT a "catch up" measure, but "exponential growth brake". With or without it, there are only 2 ways to catch up in AoS:
- Deliver longer routes than the leader can. (i.e. get to 6 level trains first. Locomotive comes into play here.)
- Prevent the leader from delivering any long routes without sacrificing your income growth. (That means you have to be able to deliver those same cubes effectively or maybe build track to block access of the leader to access those valuable cubes.)

Nicely stated!



John
Pittsburgh

In other words, there is no effective way for someone who gets behind to catch up? Getting behind in cash seems to virtually guarantee that you'll have no chance in the turn order auctions later in the game and thus, on turns when it's important, little chance of beating a leader to a key build or delivery. Maybe I'm missing something. Maybe the designers consider it desirable. (Plenty of people are happy with a game rewarding good early performance with an overwhelming chance for success at the end.)

Thanks for giving me some insight into the purpose of the mechanism, even though it doesn't sweeten the game much for me.

Randy...

Did you read Richard's post?

There are many ways for someone who is behind to catch up, some of which Richard itemises above. They require good play however, good play and a lack of later errors. There are many many examples of players losing income in the early rounds clawing back into contention with good play. It is hard certainly, but quite possible.



Then issue an extra share or two, use Turn Order Pass as appropriate (third most valuable action in the game), and get the auctions you really need. After all you have the advantage of being late in the turn order and knowing exactly how much you'll need to get what you want. Use that to your advantage -- they don't have that luxury.



A player who does very well in the early game in AoS is often likely to win, yes. This is true in many games. No guarantees tho and it isn't overwhelming. The classic case is short term income sacrifice versus long term deliveries. This is where I got waxed in this weekend's game of AoS:Wales. Richard simply timed his end-game runs better than the rest of us and had the extra Links capacity he needed to run them. A very nicely executed seemingly come-from-behind victory on his part. Ted was profitable on turn 3 while Richard and I were still paying ~$8ea in expenses and continued to be in the hole for many more turns. The difference was that Richard built his Links up faster and had the deliveries to feed it -- we didn't.

I was going to write about the comeback game--but JC beat me to it!

JC, Ted Alspach and me were playing JC's new prototype: for purposes of understanding the game there were two rules we were playing with:
- It had a way to allow longer than 6 link deliveries to be made.
- Because it was a three player game we using the "Temporary + 1" Loco rule from the Japan map.

Ted, chose to use the Locomotive to make 2 profitable runs, to boost early income and take the lead. By about turn 5 his income was about 27, mine was about 20. And he was no longer needing to take shares--while I would use my full allotment. But I had upgraded my train (permanently) once each and every turn--and could dliver 7 link routes. Ted was behind by 2 upgrades.

How'd I catch up?
Turn 6: Ted had urbanized a town to black city (he it needed badly), I was able to make a vital connection to start delivering black cubes. Ted upgraded and made only one delivery. I made 2 deliveries (in part due the black city Ted placed.) I actually took the lead in income on this turn due to Ted's upgrade--though I was still well behind since Ted took 5 fewer shares.

Turn 7: I took Urbanization, and placed a black city at the other end of my route and started a bypass--so I could deliver most of my remaining cubes for 7 or 8 links. Ted had to upgrade again. I now had a high enough income, I could beat Ted in the Auction.

Turn 8: I win the auction and get Locomotive--the pattern repeats over the final turns. I finished my bypass but needed to one link to a purple city in the south end of my line, which I started. Now I have routes for all my black cubes, With Loco I was now delivering 8's to Ted 7's.

Turn 9: I finish the route to the Purple city giving 4 key deliveries of length 8--enough to finish the rest of the game. Ted starts to run out of 7's and is forced deliver 6's. (I had taken all of the cubes in my area that Ted might have been able deliver for 6--leaving the cubes he had no interest in for later deliveries.)

In income, I went form 7 behind to 9 in front at the end. With a 5 share differential and 5 more tracks in Ted's favor--final result a win by 7 points.

The key was I had set myself up with a strong locomotive Stonger than the leader. I set up cubes for the long deliveries I'd need later in the game and made sure that they weren't useful for either Ted or JC to "steal". I made key builds to get all of the value out of my cubes.

In order to catch up you have:
- Develop your track network and improve your train. If haven't laid the groundwork, a comeback will never happen.
- Always look for your next delivery and ones after that and the ones after that. Not that you can plan everything, but if you don't plan for it, comeback won't happen.
- If you can see a delivery or track build that hurts the leader that you can take advantage of effectively (i.e hurt the leader, but don't hurt yourself too much to do it.)

All excellent comments. Thanks for the thoughtful replies. I'll look forward to taking another stab at it.

Randy...

Heh. You did well. Its also not as if this were the only example of this particular recovery art.



FWVLIW On that map it is possible to have a total across narrow and standard gauge of 9 Links, allowing for total delivery runs per turn of $18. Expenses for those links however will be $14...which can make the endgame and how far to grow your Links slightly more dicey than you might expect. It can be easy to increase Links to the point where it hurts you more than it gains you.



Remember those blacks I delivered from up north to my southern Black? Those were Ted's 7's. I ran them for 5+1 for me as I knew if I left them there they'd disappear. Ditto for a couple reds abd a blue that Ted could run for 6's -- I poached 'em for 6+1. I think this is where the game hinged and it was at around this point that I started cursing and telling y'all that you suck and ended the game on my safe 6s. I had some nice runs, but it would have cost more me to pass and bump than I could get back from the increased capacity. Your endgame long deliveries were protected. My endgame long deliveries were mostly missing and where present involved taking Ted's long deliveries. Ted, well, got caught in the middle and with me stealing from below and you hammering 8s home, couldn't keep up.

Yes, this is my gripe with the system too. I have no disagreement with the basic income reduction. However, it is artificial to have a player find that increasing his income is to his detriment - and only at these arbitrary places. My preference is to play that a player who is on the border only loses one step to income reduction - but most people prefer not to monkey with the rules... just because they are the rules. So I put up with it. But as Randy notes, it is artificial and "gamey".

Absolutely. However I have a hard time complaining about or disliking any game mechanism, anti-thematic or not, which allows players to affect each other in interesting ways, and the current income reduction rules along with the 10 cliffs certainly give players a way to affect each other in an interesting way.



That has no effect except for moving the cliff from $10 to $11. The cliff is still there, has exactly the same effect, its just $1 higher.



Yup. Moving to uniform income reduction (flat 20%, all fractions rounded down) as I've taken to doing on a few of my maps (ie the ones that I feel are a bit too rich too early) changes the net game very little, just delaying the endgame for a half turn or a turn.

AoS needs a break on income, as has been clearly stated above. I put the present version in because it seemed to work better than the system in Lancashire Railways - the LR system was a bit hit-and-miss. Effectively the AoS system said that you had rolled a 10, with a double penalty, so it's not that far removed from the LR system. I would admit that it is not an ideal solution but I have not seen anybody come up with a better way of handling the problem of reining in the leaders. Having said that I do have some ideas of a new way of dealing with the problem, but cannot say anymore now.

Martin Wallace

Martin,

I'm curious why you think the Age of Steam system works better than the Lancashire Railways system. My very, very limited Age of Steam experience suggests that the current system adds undesirable gamey-ness in that it encourages odd behaviors around the borders between income reduction areas -- people running for less that optimal amounts to stay one space under the ledge, other people running unnecessarily on them to give them one dollar to push them up onto the ledge, etc. -- and, again in my experience, seldom seems to actually tax the leaders much more than the trailers. The income bands are wide enough that I've seen the last place player only benefit once or twice in a game, far less than seems necessary to make up for the early leaders' advantages.

Perhaps, for the purposes of this discussion, you could remind us how the LR system worked?

Thanks for the games and the great comments.

Randy...

Am I misremembering the Lancashire Railways mechanism, or is this basically what it does (with the rate variable and determined by a die roll, but uniformly applied)? This seems like it would both address your concerns about exponential growth (it would be interesting to see Richard Irving redo his calculations with your 20% tax and see if it kept the hill sufficiently steep) while also satisfying my desire to see trailers less heavily taxed.

Thanks again for the great discussion.

Randy...

Roll 2 six sided dice, add 1 to the result, divide each player's score by that number. Lose that much income, rounded down.

You are misunderstanding. Moving to a flat 20% (basically $1 income reduction for every $5 of income) doesn't change the game in any significant fashion. Not a one. It doesn't significantly change Richard's charts, it doesn't change the basic challenge of exponential growth, and it doesn't change the steepness of the hill. The only notable things it does are:

a) Removes a little over $2 per player in the early game
b) Removes the income reduction cliffs at the 10 multipliers
c) Makes the timing of the 5 and 6 train slightly more sensitive if your income is near a multiple of 5

And none of those are significant. $2 in the early game is ignorable noise. The income reduction cliffs are also noise; just a minor toy for player to wield against each other and removing it changes no fundamentals. The timing of 5 and 6 Links bumps is also noise as the penalty is now effectively zero (if we had fractional income and debts it would be zero)

Income reduction is not a catch-the-leader mechanism or even really an accellerator for trailers. That's not the effective impact of the mechanism and it is disturbing to see so many continueing to describe it as such after repeated clear demonstrations (such as Richard's above) that it isn't. At best I'd describe income reduction as a way to make the game close to linear. The base idea is that it takes X effort to advance by R from Q level to Q+R level, no matter whether the starting position Q is low medium or high. That's not catch-the-leader or help-the-trailers, that's just keeping everyone playing the same game. The extent to which this attempted linearity breaks down in AoS is the degree to which players use planning to efficiently optimise X down and R up. But then that's where the skill of the game comes in.

Heh...not surprising.



Significantly, I see that it removes the tendency (undeniably present, no matter how much people may reasonably debate its effect) for tazation to punish trailers worse than leaders. A player with income of 20 asked to give up 2 income paid 10%, while a player with 11 income who also gives up 2 paid almost twice as much.



Try not to be disturbed. :-) It's just a game. And I'll agree that helping trailers or reining in leaders doesn't seem to be one of the effects of the mechanism, regardless of the designer's intent. I wish it did that a bit more.



Agreed, J.C., and I think you've done a great job explaining why that's so. Frankly, the comments that you and others have made here have renewed my interest in trying the game again, even though they don't really address some of the concerns I feel about the game. I can tell there are some good things going on in the game, I just worry that this game will wind up being like most of Martin's games for me...an almost classic. ( ...Why, Liberte, can't you be better!?!?) I'm very interested to see what the new edition brings.

Thanks again.

Randy...

Fair call, but those few dollars are really rather unimportant. You are also assuming that one player will have $11 income when another has $20 -- something that should very rarely happen. Income gaps in Age of Steam, especially that early in the game, are usually and should be around $4-$6, and probably much less.

Think of it it this way: In Age of Steam the Locomotive decision is simple: Bump every turn. That simple. Just bump Links every turn. No thought, pure automatic. (This is why some of my new maps feature a mechanism I call "accelerated expenses" which hopefully causes the automatic-bump assumption to no longer be so automatic) Assuming a fully normalised game, failing to bump Links on any turn will result in a lower final score for that player. Ergo, every player bumps every turn. The exception usually happens once for each player typically with 4 Links, sometimes 3 and more rarely (and scarily) 5, where that player fails to bump and instead delivers twice, lungeing for profitability so that they can stem the continuous share bleeding.

That $11 player you cite? He is, or should be, relatively low in the turn order. This gives him a strong advantage. He knows exactly how many shares he has to issue to get Locomotive, and as a result he can bludgeon the other players into underbidding so that he gets it cheaply. They know he's simply going to buy it, thus it is simply not worth it to them to try and outspend him. Instead their only interest is in settling the not-second player order. He then either delivers and bumps to double deliver next turn (often the best course), or double delivers that turn for $2 more income than he could otherwise have expected (very tempting but usually the wrong choice as net income is lower). The kicker is that the 2-leap in Links on the turn, followed by a double run for 2(X+2) (a total of (2(X+2)+(X+1)) from the starting turn) where X was the original Links value, should be more than enough to loft him back into contention if not the lead. Meanwhile his opponent will now be one behind in the Links race (he double-ran earlier with a smaller Links to get his income up), andis now behind on the income race. Two turns from a trailer to lead is not bad.

In many of our games the early income leader does not win the game.



Heh. I'm rather it doesn't actually, but that's probably not a surprise.



Thanks.



Cool. I'm glad I could help.



I know the feeling. For me Liberte just doesn't work as written (the draft pool clogs and other problems). Struggle of Empires needs another few months of developmental trimming even outside of the silly dice, as too much of the game comes down to who managed to crawl out of the last war most unscathed, rather than being the sum of prior turns. Etc.

J.C.,

Suffice it to say that I'm going to have to take some time to digest your discussion about the automatic nature of "deliver once and bump Links", the idea of issuing shares aggressively to get Locomotives, etc. There are certainly aspects I might not have already considered. All of this assumes some control over commodities, etc., and that's an entirely different issue, but I'm happy to try the game again, see where I might have misjudged on one issue, and then attack the problem of how terribly chumpy the Production action is. ;-)

Thanks again.

Randy...

Of course you bump every single possible turn. Bump first if you can tho, and then run for your new bigger Links. It really is that automatic. Grab a spreadsheet and do a few quick models:

1) Assume that a player bids $4 every turn, spends $7 on his track every turn, delivers for his current Links on his first delivery and then bumps his Links on his second delivery every turn.

2) Identical to #1 except this time assume that the player bumps his Links with the first delivery every turn and delivers for his new current Links on the second delivery.

3) Same as #2 except this time assume that the player gets Locomotive on his first turn.

4) Same as #3 except assume that the player gets Locomotive every turn until his Links are at 6.

5) Same as #4 except that at the first opportunity when doing a double income run would bring the player to profitability, do so.

In each case what is the value of his final income minus his final shares? See why the decision is that rotely automatic?



What else are they for?



Yeah, there's another whole game in planning track and bids so that the cubes are there for your early runs, your mid-game runs, your late game runs etc.



Heh. The actions are pretty simple really. Many Euro players seem to get hung up on the idea that the actions are not of equal value and that this is somehow a problem. It seems unfair? It isn't a problem. It is a feature and a desireable attribute of the game. Some actions are worth more than others. Locomotive is clearly the most valuable action in the game. Urbanisation et al are clearly nowhere near as valuable as Locomotive. The turn order auction is a dollar auction. The player who goes 2nd usually gets screwed: they pay way too much for something mediocre. This is a design feature of the game and why the turn order auction works so well: You want to go first, but the penalty of going 2nd is brutal. Angst!

The actions:

Locomotive is the most valuable action. That simple. It is the most valuable action. See the above spreadsheets for why. Someone wins the turn order auction, they pick Locomotive, you nod. They pick something else? You'd better perk up and figure out why something else suddenly and somehow managed to be more important than Locomotive.

Urbanisation is next -- better than Engineer because you don't have to pay for the 4th tile, and because it builds and sets delivery routes for you now and for later in the game. It can also be used offensively. All good.

Turn Order Pass is probably next (Engineer contends), simply because it will usually get you 2nd or 3rd in the turn order for free. Usually 2nd unless the other players are working against you. That means a free Urbanisation action (and sometimes Locomotive!) -- better than paying twice for two things not nearly as good -- but you'll have to plan two turns ahead.

Engineer is the next least valuable simply because you have to pay for the extra track. Sometimes necessary tho.

Production is next. A nice little action, especially from the mid-game onward, nothing special, but often enough a game saver or breaker. Useless the first and last turns of course.

Both First Build and First Move are basically wastes of money. In general they should only be chosen if you really are at serious risk, or you're late in the turn order (ie paid nothing anyway), and can use them to reduce the value a player who spent a lot will realise from their earlier turn placement (always a Good Thing). First Build is a popular choice for the 3rd or 4th player on the first turn for obvious reasons.

With moderately rare exception you should always see the actions chosen in the above order. This is /good/.

Hmm? It has the obvious effect of fixing the thing Randy and Jonathan were complaining about, i.e. in the standard rules, sometimes your income ends up worse because of it being increased, which is surprising and feels wrong and "gamey" to many people. It breaks the natural model people have of "increasing my income is always a good thing" and adds a wacky "except at every 10th value" exception.


It may have little effect on average overall, but it certainly has a concrete effect of removing a "gamey" effect that jars many people and/or takes getting used to...

That said, I can respect your argument elsewhere that it provides a weird interesting way for players to mess with each other.