Forum changes: Editing of posts has been turned off until further notice.

Main Menu

no islands in the pool

Started by Paul Czege, September 25, 2001, 09:32:00 PM

Previous topic - Next topic

James V. West

The Anti-Pool flip-flop idea is interesting. What we're dealing with is this: The Pool stresses risk, Anti-Pool stresses balance. I can see both working well, but with different feels. Obviously I like the feel of the risk factor. That's why I designed it that way. Its my preference. I don't like games in which I don't feel like there's something to be lost. However, I don't like games in which you feel like you just can't win. I don't want The Pool to be one of those games.

Eloran mentioned making Traits cheaper. I don't think that would solve this problem. If you did that, and each player then started the game with more dice but still kept the same amount from game to game, the problem would persist. I do see your point that higher bonuses in some Traits would make it considerably easier for players suffering bad luck. Good point. But the cost system has to be intuitive. No charts, no complexities. Either cost equals bonus (too cheap) or cost equals bonus times bonus (expensive, yes). For now, I'm with keeping with the current cost system.

I've been thinking about this problem non-stop since Paul emailed me about it. Maybe the solution isn't as complicated as I first thought it was going to be.

The problem lies in players' pools getting emptied too quickly. When you create a character, you get 12 dice and then you spend some of those and you use the remaining dice in your first session. Ok, that leaves an average of about 5 dice for a starting character. Five dice for an entire session of play obviously isn't going to cut the mustard.  It needs to be more.

I think of each session as a chapter in a story. In each chapter, characters do things and some of those things are more important or memorable than others. So it makes sense that you'd have a finite pool to draw from and that at the end of a chapter you might be broke. Maybe you went out with a bang, or you went out with a floop. The problem that Paul has expressed is that at the beginning of each session his players already have little or no dice, then with the 1d6 roll that only boosts them about 3 or 4. So they're probably starting a game with 4 dice in their pools. That sucks.

I think that at the beginning of each session, just like a new chapter of a book, the pools get reset. Each player gets to start with a fresh and generous pool of effectiveness and they can then make decisions about how to gamble those dice throughout the chapter. This would work best if the GM were actually thinking in terms of chapters, resetting the pools after each chapter has finished. Perhaps more than once per session depending on how the story is being paced.

It seems to me that this is a viable solution to the problem. New chapter, new pool.

How many dice to reset to? Perhaps this is a question that needs to be answered by each individual group. I can offer suggestions based on experience, but that experience is still being developed so right now I'd have to throw out some arbitrary numbers based on intuition:

12 dice gives you the ability to make the max gamble of 9, and still have a few left if you fail. But it keeps the pool small.

12 plus 1d6 gives a slight boost and a certain amount of randomness and anticipation at the start of the game.

15 dice is a generous number that might make players more apt to gamble early.

I'm a bit leery of starting higher than 12, but I'm keen on the idea of getting 12 plus some.

Either way, this process lets players who have suffered bad luck get back on track in shorter time but doesn't reduce the risk of bad luck too much to render the risk factor impotent. Make sense?

I'd really love to hear your opinions on this. When I designed the game, I was keeping the figures low intentionally and I knew that I might have to do some tweaking later. Now I'm tweaking.

James V West

James V. West

Paul mentioned a couple of serious potential flaws in the idea I mentioned above:

1) Not every group will play "chapters" the same way. Some chapters will be more intense and require/induce more gambling while others will be less game-intensive and more drama-intensive, thus calling for less rolling.

2) Players may fall into a habit of waiting for a cue from the GM as to when a chapter will end and then using dice like mad, casting a kind of "meta-gameish" shadow over the story.

Good points, Paul. I'm not sure yet if this will shoot my idea out of the water, but it might.

James V West

Mike Holmes

James and all,

OK, first off, I think that The Pool as written is a fine game, and that's the last time I'll say it as I've now put it in several threads. First off your experience, James, has been that players do not bottom out consistently. And even when they do, as Scott Knipe attests in the Metagame resource thread, the game still contiues to work. It just takes on some different characteristics. Which might, for all I know, be the best way to play.

FWIW, the risk in the Anti-Pool is in trying to succeed while rolling low amounts of dice so that you don't deplete your pool. You have an incentive to do so in that failure means that you get dice. So, yes the risk is more balanced, but it still exists. In regular The Pool, there is a best choice given the number of dice available. For example, with a large number of dice rolling them all makes the most sense whether or not the task is important from a gambling perspective. Unless you don't mind failing and losing dice.

Interestingly, it is true of all gambling on something with a fixed number of results that there is always a best strategy. So when playing craps, or blackjack, etc. there is a best way to play. In casinos playing the best way will end up with you losing about 1% or so of your money bet over time on the average (depending on the exact rules in play). This is how casinos make money. In fact with blackjack, you can use a best practice of counting cards that can over a very long time make you money (under the best circumstances about 2%) assuming that you don't get thrown out first, which you will unless you are Dustin Hoffman.

Often people are unaware that there is a best method to playing these games and on the average they lose their money much faster, many believeing that by going with premonitions they can do better than any statistical strategy. This is not supported scientifically, however, and the casinos do very well, you'll notice. This is one thing that may cause people to not use the best strategy in The Pool, in addition to the facts that some victories are more important than others, some people might want to fail, and the fact that there are methods to gain dice automatically (something you can't do in a cassino). And also that losing dice just leads to a different story, not actual financial loss. But to that extent, then we are dismissing the idea of it being a gamble. If you want the pressure, you have to accept players looking for the best strategy.

And if I hadn't blurted it out, Sullivan or somebody else would have at some point. :razz:

The balance (the consolation prize) in Anti-Pool makes it so that any level of expenditure may be acceptable given a level of importance of the roll. So players will make different choices more often based on the needs of the story, rather than the best gambling strategy. This may appeal to some. What it does is put the player back in more control (or at least prevents him from having conflicting priorities), and paces the game that way. I posited it pretty much soley to discuss other ways that such a mechanic could be constructed, and what effects doing different things with it might have. I've been very pleased with the responses so far.

The Pool, OTOH (and I think possibly much more interestingly), by giving the player conflicting priorities drives the story along on it's own cycle whaen the player gambles. Note that he can refuse to gamble much and take more control. If the cycle it generates, however, turns out to be a good one, then I wouldn't change anything. You'll have found a way for a metagame mechanic to pace a story for in a semi-random manner. That's really a cool idea. This is essentially what I'm looking to discover in the Metagame Resources thread.

The only doubt I have about The Pool's ability to pace really effectively is that I don't think that you did any really careful analysis before deciding on the particualar numbers or methods and just went on instinct. Which I think has served you well, but might possibly be tweaked. This is why I'm excited to see you looking into it more closely. Even just discussing possibilities may lead to a better understanding of the phenomenon.

OK, that was to wordy. To summarize, The Pool is cool, and so is Anti-Pool and other discussons arising from it. Howzat?

Member of Indie Netgaming
-Get your indie game fix online.

Mike Holmes

On 2001-09-27 18:24, James V. West wrote:
Paul mentioned a couple of serious potential flaws in the idea I mentioned above:

1) Not every group will play "chapters" the same way. Some chapters will be more intense and require/induce more gambling while others will be less game-intensive and more drama-intensive, thus calling for less rolling.
This could be a good thing. It might mean that short chapters and long ones will tend to have different feels to them. Might be bad, though, too especially in the case of very short chapters where the player's pools never even have a chance to dwindle. Still, even that might be interesting.

2) Players may fall into a habit of waiting for a cue from the GM as to when a chapter will end and then using dice like mad, casting a kind of "meta-gameish" shadow over the story.
Well, the gambling part already does that. This might actually be a good thing as it will promote thrilling climaxes. And the potential for false climaxes wold be interesting. You could fake the group out and get them in your deathtrap after they deplete themselves. Then you go to a cliffhanger and next week they replenish and have the dice they need to escape. Cool.

Member of Indie Netgaming
-Get your indie game fix online.

James V. West

Thanks for the analysysyis (anilysisiys?) Mike.

Yeah, The Pool was pure gut instinct. I don't like calculations.

Upon further thought, the only real change I'm considering making at this point is to allow players to roll 2d6 instead of 1d6 at the start of each chapter, that being added to their total carried over from the previous chapter. Not a huge change, but it's certain to give players more dice on average.

The other ideas I've come up with all suck. I don't like any of them for one simple reason: they aren't integrated. Even the idea I just mentioned isn't really what I want. I'm not convinced that there needs to be any major alteration, but if I make one, it will be something fully integrated into the existing structure.

So for now, I'm sailing on with my game design and I'll see what develops in time. I'm playing again Monday night. I'm planning a session that ought to induce a lot more die rolling and thus more gambling. We'll see what happens :wink:.


James V. West

Paul Czege

Hey Mike,

Total--Chance of

How difficult would it be to do the same math for a Pool variant that used four-siders instead of six-siders?

My Life with Master knows codependence.
And if you're doing anything with your Acts of Evil ashcan license, of course I'm curious and would love to hear about your plans

Mike Holmes

Member of Indie Netgaming
-Get your indie game fix online.

Mike Holmes

And before anybody asks, here's the same for D8 and D10,


Member of Indie Netgaming
-Get your indie game fix online.

Paul Czege

That's food for thought. Thanks.
My Life with Master knows codependence.
And if you're doing anything with your Acts of Evil ashcan license, of course I'm curious and would love to hear about your plans

Mike Holmes

Actually, the formula is simple for anyone who wants to calculate any other die or any number of them.

P = 1-((n-1)/n)^d

This returns the probablity of success "P" where "n" is the highest number rollable on the die, and "d" is the number of dice rolled. For those not familiar with programming, "^" means "raised to the power of".

The ((n-1)/n)^d part alone is the chance of failure.


(edited because I originally posted the formula as (n/(n-1))^d which is incorrect, a fact that I noticed after I saw Mike do it right below. Whoops, going too fast! Sorry for any inconvenience.)

[ This Message was edited by: Mike Holmes on 2001-10-05 17:23 ]
Member of Indie Netgaming
-Get your indie game fix online.

James V. West

Hey Mike, I bet all the kids made fun of you in math class. "hey look, there's calculater boy!". :wink: Just ribbing. Thanks for the breakdowns, but how do you figure in the fact that you're rolling several dice and not just one? I mean, it seems to me that if you were rolling percentile dice, then that's just one roll. But rolling a handful of dice gives you a handful of opportunities for different results, doesn't it? Maybe that's not mathematically sound thinking, but it makes good horse sense.

I like the odds, by the way. Even a big ole handful of dice doesn't mean a guarantee of success, while a small handful doesn't mean definite failure. Just the way I like it.

D4s represent both the beauty and the evil of dice. Four-siders are clumsy. They have two redeeming qualities: 1) they give you quarter results and 2) they look like pyramids. But holding a handful and rolling them sucks.

Now, there are many ways to tweak a game. I *could* do any one or several of the following:

1) Build the game around blank d6s with dots on only two sides. That increases your odds by a factor of 1/6.

2) Say that a one or a six are both successes. Then add some funky rule that says you can only get a MoV on a one.

3) Give 2 dice per success instead of 1, thus doubling the flow of dice into the pool without adding any goofy new rules.

4) Add dice to the pool based on traits used in the last game.

And many, many more.

The crux of the problem for any chnage in the rules is this: there can be no scaling.

Scaling, as in counting successes for example, means that invariably one style of playing is going to prove more "successful" than another. For another example, say I create a better dice-refilling mechanic by saying you get dice equal to the bonus of the Trait you used in the roll. Well, automatically players will tend to use one or two Traits as often as possible and ignore opportunities to use their Traits with smaller bonuses (or none at all). No good.

Say I reverse that, and say that you get something like 4 dice per success *minus* your Trait bonus. Well, in that case, no one would ever add a bonus to a Trait. Period.

So, there can be no scaling. It has to be a flat black and white mechanic. YOu roll a one, you succeed. YOu don't you fail. That's it. You succeed you get a die or a MoV. That's it.

Now, I realize that by looking at it statistically you can argue that betting high every time is a better way to go. All I know is this: I've ran three sessions, no one ever bet more than 3 dice on any roll and no one ever lost all their dice. Why? A couple of reasons.

a) Not a lot of dice rolling. There was a lot of drama going on and the players felt they didn't need to do any gambling in most cases. But when the tension mounted, they threw the dice. Rates of failure and success were about even in most cases, adding to the game's drama.

b) No excessive hosing. If someone bet a bunch of dice and lost, sure, the failure should be significant. But if they only have one or two dice and they lose them, there's no real need for a serious hosing. Just call it a failure. The way I typically handle failures is to go with the mood of the scene. If I feel like the story would be better served to say the failure was a kind of success wrapped in a failed attempt, then so be it. Maybe someone was trying to do one thing, which certainly failed, but casused something beneficial to happen at the same time. Its a story game first and foremost.

In fact, on the issue of failures...

When I run this game, I find myself thinking that "failure" isn't the right word for a failed roll. When no dice are being cast, its simple drama and storytelling, right? When you roll you're prize is to be successful and to get a die or MoV. A failed roll usually results in losing some dice and that's bad enough. So, I often tend to treat failed rolls as a chance for me, the little ole GM, to let the event play out as *I* would like it to. That doesn't mean kill the PCs, it means make a cool scene.

And it isn't just with The Pool that I do this. I've always done this even with the most anal retentive games I've ever played (like GURPS).

I'm paying very close attention to the language I'm using in writing my game. The current version of The Pool is not perfect. It needs revision not in rules, but in language. Once I've finished writing The Questing Beast and I think the rules sound the way they are intended to be used, I'll put up a revised and more insightful (hopefully) version of The Pool.

You guys have been extremely helpful in all this. Thanks!

James V. West

P.S. and thanks for letting me ramble on



Essentially, here's the logic behind Mike's formula.

Your odds of success are equal to your odds of not failing, right?  And it's pretty clear that, with a single d6, your odds of not failing are 5/6, right?

Okay, now suppose you have a lot of d6's.  Instead of rolling them all at once, let's suppose that you rolled them one at a time.

So, you get the first roll.  It could be a 1, in which case we don't care what the rest of the dice roll, or it could be a 2-6, in which case we do.  So, the second roll could be 1 through 6 for each face of the first roll (that is, you could roll a 3 on the first die and a 2 on the second, and that's a different roll from 3 on the first die and 4 on the second).  We count up all the failures -- there's a 5/6 chance of failure on the second die for every (failing) face on the first die -- or (5/6) * (5/6).

Add a third die to that, and you've got a 5/6 chance of failure for every combination of the first die, or (5/6) * (5/6) * (5/6).

Obviously, there's a pattern, here.  If you roll a series of dice, your odds of failure are (5/6) ^ n, where n is the number of dice you rolled.

If you use 1 - (5/6) ^ n, that's your chance of success, since everything that's not a failure is a success.

Now, the final thing we realize is that it doesn't really matter whether we roll them in order or not.  The dice don't relate to one another in any way -- so we could arbitrarily say, "that one's the 'first' die, that one's the second," etc. even if we roll them all at once.  Since there's no difference between rolling them in sequence or all at once, the (5/6) ^ n formula must also be the chance of rolling a failure with a bunch of dice thrown at once.

Mike Holmes

Whoops. First of all I made a mistake in my formula. The stats I posted are correct, just not the formula. It is now corrected and anotated. So, if that caused any confusion, I apollogise.

Secondly, I've never advocated either hosings based on level of expenditures (although I did argue with somebody that, if you did that, you should instead base it on the number of dice bet for several reasons). I also have never advocated making the level of success based on the number of ones rolled. I'm not against a system producing such figures, however, I agree with James that, in combination with this particular system, you may get problems.

What you suggest James is essentially a D3. Work out the stats yourself to see what effect this has, or just check the difference between D6 and D4 to see where the trend takes you. Interestingly, there are a few games that use D2 (AKA coins) to produce the same kind of effects. In general, it would seem like you'd need more if you used these. I think that the bigest advantage of D6's is that they are common. Anyhow, changing dice type just shifts the curve without really changing any of the dynamics. The same strategies hold for any sized die in this case.

Which is why I wonder what Paul is up to?!?

In any case, you'll notice that as the die pool gets bigger you get diminishing returns, FWIW. Each further die means less in terms of success (but are still worthwhile strategy-wise). Just a note.

Member of Indie Netgaming
-Get your indie game fix online.

Blake Hutchins


Something you mentioned earlier has me wondering whether I understand The Pool correctly. You said starting with 12 dice would allow a player "to make the max bet of 9." Is this a rule you're contemplating, or did it come up earlier in this thread (in which case I've simply missed it)? I don't see anything in The Pool about a limit on the number of dice one can gamble.



James V. West


Nine is the limit of dice you can gamble from your pool on a single roll. This doesn't count the dice you get from the GM and from Trait bonuses. Its in the rules, but I think I only mention it once.

I actually put nine dice in my hand and felt like it was a comfortable maximum. Any more and you can't do a one-handed roll. Plus nine just felt like a natural limit, so I went with it.


James V. West