Who the hell comes up with the time to play a scenario on this site?

[Psycho]

But how about for a fair to middling player like me Grumpy Fort? You know how slow I am coming from my Mississippi public school background.

You people from Mississippi sure is not smart!

 

[Turuk]

I tried to correct the ROAR results in the Archive, but it isn't possible because AP206 is not yet in ROAR, like many other scenarios.

I am checking with Dave on functionality to set ROAR to a blank value as I think this may happen more in the future as the Archive and ROAR grow more out of sync with new content (and scenario names overlap).

 

[Larry]

I would venture to guess that play time correlates to several factors -- experience, how bad the player wants to win, situational familiarity, etc.

https://en.wikipedia.org/wiki/List_of_world_records_in_chess#:~:text=The longest game played in,15 minutes and 40 seconds. Longest game time for a world championship chess match over 8 hours.

 

[daveramsey]

Some of the scenarios from AP19 have absolutely absurd times to play numbers.
What sort of method is used to determine these times...Dartboards?

And, there is a scenario that has no playings, yet has a balance number listed?
Really quite off-putting to say the least.

GF

Hey Gary, I'll assume this was a couple of well-meaning questions so I'll answer as informatively as I can!

The archive uses a formula that is generally ok for many scenarios but is stretched by factors that it doesn't know. So, for example, reinforcements that come on late in the day, scenarios with little movement, or complex rules.

The exact formula is as follows: minutes = (total infantry squads * 0.7) + (total guns * 0.6) + (total afvs * 1.2) * average number of turns (to a minimum of 5) * 2

If there is oba for one side 30 mins is added, and if both sides have oba 1 hour is added. Night adds 1.5 additional hours.

I have the ability to override individual scenario timings, so it's possible that some don't adhere to this formula, but this are very rare, and it's often done because the scenario is strange in some way (lots of turns or a tiny map, etc).

In terms of scenarios with no playing but have balance - as turuk pointed out, sometimes scenarios have the same names as previously logged scenarios in roar from different publications. This works well when scenarios are republished, but breaks when the names match for completely different scenarios. I don't have a clean fix for this now, other than to wait for roar to include the new scenarios in the database and then I'll link them to the new one in roar. By all means provide examples and I can see what I can do for these scenarios.

Sorry if you've struggled with these, I'm happy to look and improve things when I get feedback that I can act on.

Cheers,

Dave

 

[Fort]

Hey Gary, I'll assume this was a couple of well-meaning questions so I'll answer as informatively as I can!

The archive uses a formula that is generally ok for many scenarios but is stretched by factors that it doesn't know. So, for example, reinforcements that come on late in the day, scenarios with little movement, or complex rules.

The exact formula is as follows: minutes = (total infantry squads * 0.7) + (total guns * 0.6) + (total afvs * 1.2) * average number of turns (to a minimum of 5) * 2

If there is oba for one side 30 mins is added, and if both sides have oba 1 hour is added. Night adds 1.5 additional hours.

I have the ability to override individual scenario timings, so it's possible that some don't adhere to this formula, but this are very rare, and it's often done because the scenario is strange in some way (lots of turns or a tiny map, etc).

In terms of scenarios with no playing but have balance - as turuk pointed out, sometimes scenarios have the same names as previously logged scenarios in roar from different publications. This works well when scenarios are republished, but breaks when the names match for completely different scenarios. I don't have a clean fix for this now, other than to wait for roar to include the new scenarios in the database and then I'll link them to the new one in roar. By all means provide examples and I can see what I can do for these scenarios.

Sorry if you've struggled with these, I'm happy to look and improve things when I get feedback that I can act on.

Cheers,

Dave

Thanks for the reply, Dave.
I have come to believe that a 'time to play' formula is inherently flawed and a disservice to the ASL public.
Such an arbitrary device can, and does, make players shy away from 'too long' notions based on a very subjective formula pulled from essentially thin air.

G.

 

[daveramsey]

It was a sought after number for tourney organisers who were using it to get an idea of scenarios to select for various rounds. Players, I think, are a bigger factor than the formula itself.

Who knew you can't please all of the people all of the time?

 

[Fort]

It was a sought after number for tourney organisers who were using it to get an idea of scenarios to select for various rounds. Players, I think, are a bigger factor than the formula itself.

Who knew you can't please all of the people all of the time?

Tournament organizers are using an arbitrarily derived value to pick scenarios for their tournaments?

Why not just use a dart board?

I agree that individual players are definitely a bigger factor in determining time to play.

You state in your first post that it is 'exceedingly rare' that the formula proves inadequate or just plain wrong. I would like to see the data on that claim. My experience has been the values given by this method are way off, sometimes by as much as a 3-4 hours overestimate.

I believe this formula is a disservice to players, designers and tournament organizers.

 

[Jeffrey D Myers]

Dave's response to "Ravioli Ravioli Give Me the Formuoli" is met with a ton of gravitas....

 

[Stewart]

I am a very experienced player and the times for the scenarios in AP19 are way off. The times to play are inflated by quite a bit.

A very experienced player will play any of those scenarios in a tournament round of 4-5 hours with time to spare.

How long will it take for an Average player to play those games?

Some of us aren't in a RUSH to CHESS CLOCK our games.

Avg player 1hr per turn...Considering many players simply give up too early....less time.
I'd love to know how often players Concede games....

 

[Fort]

Depends on the player. I double the time and it's about right for how slow we play. I wouldn't put much stock in any estimated play time.

Thank you for making my point for me.

The formula is useless as a tool for determining time to play any scenario. All it does is discourage the inclusion of scenarios that don’t readily fit its mold.

 

[Fort]

How long will it take for an Average player to play those games?

Some of us aren't in a RUSH to CHESS CLOCK our games.

Avg player 1hr per turn...Considering many players simply give up too early....less time.
I'd love to know how often players Concede games....

Here again is an arbitrary number thrown out there…”average player 1 hour per turn.“

Where is your data?
Are you an ‘average’ player?
How many ‘average’ players have you observed?
What method did you use to time and record the progression of the game?
Is every turn 1hr per turn from first to last turn?

 

[daveramsey]

Tournament organizers are using an arbitrarily derived value to pick scenarios for their tournaments?

Why not just use a dart board?

You can use the same argument for balance.

But, back to the time to play. It's clearly not an arbitrarily derived value. It's a formula. You can argue against the formula's composition, but suggesting it is essentially random is clearly disingenuous. We know that players' speed is on a spectrum, and that spectrum is exacerbated by rules complexity, platform, environment and force-composition. Each of which the archive doesn't have full visibility over.

You state in your first post that it is 'exceedingly rare' that the formula proves inadequate or just plain wrong. I would like to see the data on that claim. My experience has been the values given by this method are way off, sometimes by as much as a 3-4 hours overestimate.

I believe this formula is a disservice to players, designers and tournament organizers.

That's a bizarrely combative statement. I can't give you data on something that can't be measured - you've already seen in this thread that some people think some values are too low, some too high. The archive attempts to provide a quick estimate on playtime, based on nothing but a top level idea of the OoB and a couple of variables. It's an attempt to be helpful, and generally speaking these values are ok for most people.

I can go through my emails to dig up the ones where organisers have asked me to add the value to various reports if you insist, and we have to assume that they're the guys that feel there's some value there. Most often these are used as a guide before they check it out themselves. If you want to be a constructive force in this area, feel free to suggest a more appropriate formula and I can look at incorporating that instead. You could also be helpful and provide some data of your own playing time from a number of different sized scenarios, so I know where you're coming from.

 

[Sparafucil3]

How long will it take for an Average player to play those games?

Some of us aren't in a RUSH to CHESS CLOCK our games.

Avg player 1hr per turn...Considering many players simply give up too early....less time.

I think Fort's main consideration is seeing his work played. The likeliest place for that to happen is at a tournament. If you assume the typical tournament has room for about 6-7 hours of play per round, anything "longer" that 5 hours or so would get thrown out without consideration.

If players were more honest with themselves about how they play, they should be able to look at a scenario and tell how long it would take them to play it. The first question I ask when I sit down to play is "how fast do you play because it will help us decide what to play".

I'd love to know how often players Concede games....

I bet it is way more often than you think. -- jim

 

[Sparafucil3]

We know that players' speed is on a spectrum, and that spectrum is exacerbated by rules complexity, platform, environment and force-composition. Each of which the archive doesn't have full visibility over.

Maybe instead of a direct "this will take 7.5 hours", you instead say "7.5 hours +/- <some percentage of your estimate>". That "fudge-factor" would account for player speed.

That's a bizarrely combative statement. I can't give you data on something that can't be measured - you've already seen in this thread that some people think some values are too low, some too high. The archive attempts to provide a quick estimate on playtime, based on nothing but a top level idea of the OoB and a couple of variables. It's an attempt to be helpful, and generally speaking these values are ok for most people.

Consider adding a "length of play" field and "turn of conclusion" field to your data collection. You could display an average actual play time field and use the turn of conclusion to adjust actual playtime if you wanted. This would probably be more accurate given enough playings. You could then, over time, modify your formula to more accurately reflect the data gathered.

I can go through my emails to dig up the ones where organisers have asked me to add the value to various reports if you insist, and we have to assume that they're the guys that feel there's some value there. Most often these are used as a guide before they check it out themselves. If you want to be a constructive force in this area, feel free to suggest a more appropriate formula and I can look at incorporating that instead. You could also be helpful and provide some data of your own playing time from a number of different sized scenarios, so I know where you're coming from.

I think Fort's point is if tournament organizers are looking at your values to make decisions, then some of his scenarios are thrown out without consideration. IMO, changing to a range as I suggest above might mitigate some of that. There is a huge psychological difference between "This will take 7.5 hours to play" and "This will take 7.5 hours +/- 1.5 hours". For one, the second has the possibility of fitting into the round. JMO, YMMV. I hope this is constructive. -- jim

 

[Fort]

You can use the same argument for balance.

Balance is a crap shoot. The most important factor in balance is player skill disparity. Most players at least look at a scenario to get a feel for the balance. With the 'playing time' declaration, a TD using this as a method for choosing scenarios will be doing a disservice to the attendees as perfectly viable scenarios are going to be overlooked.

But, back to the time to play. It's clearly not an arbitrarily derived value. It's a formula.

Ok, so it's an arbitrarily constructed formula that results in suspect values.

You can argue against the formula's composition, but suggesting it is essentially random is clearly disingenuous. We know that players' speed is on a spectrum, and that spectrum is exacerbated by rules complexity, platform, environment and force-composition. Each of which the archive doesn't have full visibility over.


That's a bizarrely combative statement. I can't give you data on something that can't be measured - you've already seen in this thread that some people think some values are too low, some too high. The archive attempts to provide a quick estimate on playtime, based on nothing but a top level idea of the OoB and a couple of variables. It's an attempt to be helpful, and generally speaking these values are ok for most people.

If you can't measure the data, then how can you construct a formula and make a proclamation that your values and results are the correct values and results?

I can go through my emails to dig up the ones where organisers have asked me to add the value to various reports if you insist, and we have to assume that they're the guys that feel there's some value there. Most often these are used as a guide before they check it out themselves. If you want to be a constructive force in this area, feel free to suggest a more appropriate formula and I can look at incorporating that instead. You could also be helpful and provide some data of your own playing time from a number of different sized scenarios, so I know where you're coming from.

To your last point, I do not believe there is a formula that adequately and fairly gives a good estimate on time to play.
I could add data from my playings of scenarios, however it would be useless to anyone but me, and my opponent at the time, as these times a HIGHLY subjective.

 

[Fort]

I think Fort's main consideration is seeing his work played. The likeliest place for that to happen is at a tournament. If you assume the typical tournament has room for about 6-7 hours of play per round, anything "longer" that 5 hours or so would get thrown out without consideration.

If players were more honest with themselves about how they play, they should be able to look at a scenario and tell how long it would take them to play it. The first question I ask when I sit down to play is "how fast do you play because it will help us decide what to play".


I bet it is way more often than you think. -- jim

Thanks, Jim.

Agree on all points.

 

[daveramsey]

Yeah - look I agree that it's not perfect and that people should consider there to be an assumed fudge factor to make it work. It gets a bit messy on the UI if I start displaying "+/- 1.5 hours" or whatnot, when I assume most people know these times are literally (as written!) estimates, rather than a specific time to play.

For scenarios which most of us can agree look out, I absolutely can and will reduce the time, but so far, I've not seen any specific examples. Just let me know and assuming these times feel reasonable to the average player (I know, two subjective measurements...) I can update them.

 

[Sparafucil3]

Yeah - look I agree that it's not perfect and that people should consider there to be an assumed fudge factor to make it work. It gets a bit messy on the UI if I start displaying "+/- 1.5 hours" or whatnot, when I assume most people know these times are literally (as written!) estimates, rather than a specific time to play.

I defer to you Dave. Your site, your decisions. I will say there is a psychology at play here. The more decimal places you display, the more accurate your estimate feels. This feeling of accuracy undermines the variation you acknowledge existing in the estimate. It is a tough thing you're trying to do here. Best of luck threading this needle. -- jim

 

[Actionjick]

At events where the TD provides a list of scenarios to choose from has the TD actually played all those scenarios?

Personally I think there are too many variables to make any attempt at determining scenario playing time for the general population of players anything but a WAG.

 

[Sparafucil3]

At events where the TD provides a list of scenarios to choose from has the TD actually played all those scenarios?

That depends on the tournament. I know Albany plays all (or almost all) before they go on the list. I know others have the same approach but I cannot speak with authority. I know Dave Goldman plays a LOT of ASL which I think feeds the playlist for the Chicago ASL tournament. -- jim