2020 cards

They keep changing card prices. A bunch of pitchers are priced at $5, which looks like a place holder. Trout's price went up. I don't think they have many resources committed to 365, especially the pricing of the card set. Who does the pricing, and how are prices determined?

They've only had six months to figure it out. Give them time

Thanks for the memories SOM. There are other games in town. I play a couple of leagues of Strat baseball, but we start our annual gatherings to draft, talk baseball, etc. early March. 365 release is too late this year for it to be included, we foolishly assumed the new season would be out by now, but after no responses from SOM, we're better off assuming 365 is no longer a priority for them, and therefore we have to move on. Hope everyone enjoys the season.

another 1-2 weeks folks........Lost Vegas is never wrong lol

The man who does the card pricing is Mark (childsmwc, or bbrool at one time I believe).

He did not have six months. The new cards were just released a short time ago, and I don’t think everyone understands the amount of effort that he personally puts into card pricing.

I played with Mark in a few leagues, and when a discussion of card pricing broke out he wrote the following:

childsmwc wrote:
For those of you that don't know me, I have helped in pricing the cards for a significant number of years now. The model currently used for pricing is one I developed as a user of the game to evaluate the card set. I have tried on occasion to start public threads to discuss pricing but they quickly devolve and become useless.

So I thought I would start some threads here, explain how the cards are currently priced and will gladly answer any questions. I am always looking for good feedback that can be used to upgrade pricing.

The first step in pricing the cards is to convert the raw data into an output that can be measured, in this case runs. I use the linear weight formula inspired by Paul Johnson called estimated runs produced. I might have tweaked the values over the years, but here is the basics:

BB & HBP=.32
Hit=.1589
TB=.32
SB=.19
Out= (.0979)

In valuing a card I do not use all of the details on the card, I use only the details provided in SOM's data disk. Therefore I do not differentiate between fly ball, K, etc. when it comes to outs. Double plays are a different matter.

Derived results:

While most results on the card are always the same result, certain card results are contingent on the situation (i.e. clutch, double plays, stolen bases) or on the ball park for the triangles and diamonds on the cards.

Also in deriving results I also make an estimate for the average pitcher, since 50% of the rolls come from the pitchers card. The reason I do this is for three primary reasons:

1)Impacts of weak on the hitters card changes the pitchers card results so the assumption you make about the average pitcher impacts this value

2)I have the ability to change the average pitcher based on the match up, so I can weight a RH pitcher tougher against righties than lefties, etc. This inherently builds in some added value for switch hitters

3) Forecasting stolen bases- stolen bases is a function of first getting on base, so I find adding in the pitchers card helps refine the estimate.

Here are the assumptions made in the model to derive situational results:

Clutch- I assume that only 10% of the time will the clutch situation be present (i.e. two outs with a runner in scoring position) when a clutch roll is made. I convert outs to singles or singles to outs based on this probability times the clutch probability. Also jumping ahead a little, I also value the clutch event differently for runs value since it is situational, but more on that later.

HR and Single park events- I convert these to singles, homeruns, and outs based on a 9/9 environment.

double plays- I convert ground ball a's assuming 20%, result in a double play

Stolen Bases- I am not going to fully describe this process, but it is based on OBP less HR's, less Xtra base hits. This result is then filtered through the lead probability's from the card and their safe chances to determine how often they will be sent, to arrive at how many SB chances a player will have. Depending on the steal chance, I multiply the SB chances times a success rate, 65% is the minimum I use, below that and I assume the runner will not attempt the steal. Note I currently do not penalize runners further for have caught stealing results while attempting the lead. There is obviously an impact for this but the effort verse the overall value wasn't worth the work to develop.

Average pitchers card

The key assumption here is that I assume 2 ballpark homeruns and 2 home run events on the pitchers card and a few OBP points and Total bases different depending on the hand of the pitcher

So lets stop their for a minute. All of this so far is just converting a players own card into Runs. Additionally I will ultimately add on:

Defense
Run rating (a function of run rating and OBP)
Team OBP (NERP concept)
Potential platoon buff if one side of a players card produces a run value below that of a $.50 cent player

I do not value hit and run or bunting into a players card

Mark

I originally put this in The Big Post for 2016, you can read more there if you like.

Is he doing this on a yellow legal pad with a pencil? Honest question? I actually assumed cards were determined by algorithm's like everything else in the world today. Is every card developed by a person?

The man who does the card pricing is Mark (childsmwc, or bbrool at one time I believe).

He did not have six months. The new cards were just released a short time ago, and I don’t think everyone understands the amount of effort that he personally puts into card pricing.

I played with Mark in a few leagues, and when a discussion of card pricing broke out he wrote the following:

childsmwc wrote:
For those of you that don't know me, I have helped in pricing the cards for a significant number of years now. The model currently used for pricing is one I developed as a user of the game to evaluate the card set. I have tried on occasion to start public threads to discuss pricing but they quickly devolve and become useless.

So I thought I would start some threads here, explain how the cards are currently priced and will gladly answer any questions. I am always looking for good feedback that can be used to upgrade pricing.

The first step in pricing the cards is to convert the raw data into an output that can be measured, in this case runs. I use the linear weight formula inspired by Paul Johnson called estimated runs produced. I might have tweaked the values over the years, but here is the basics:

BB & HBP=.32
Hit=.1589
TB=.32
SB=.19
Out= (.0979)

In valuing a card I do not use all of the details on the card, I use only the details provided in SOM's data disk. Therefore I do not differentiate between fly ball, K, etc. when it comes to outs. Double plays are a different matter.

Derived results:

While most results on the card are always the same result, certain card results are contingent on the situation (i.e. clutch, double plays, stolen bases) or on the ball park for the triangles and diamonds on the cards.

Also in deriving results I also make an estimate for the average pitcher, since 50% of the rolls come from the pitchers card. The reason I do this is for three primary reasons:

1)Impacts of weak on the hitters card changes the pitchers card results so the assumption you make about the average pitcher impacts this value

2)I have the ability to change the average pitcher based on the match up, so I can weight a RH pitcher tougher against righties than lefties, etc. This inherently builds in some added value for switch hitters

3) Forecasting stolen bases- stolen bases is a function of first getting on base, so I find adding in the pitchers card helps refine the estimate.

Here are the assumptions made in the model to derive situational results:

Clutch- I assume that only 10% of the time will the clutch situation be present (i.e. two outs with a runner in scoring position) when a clutch roll is made. I convert outs to singles or singles to outs based on this probability times the clutch probability. Also jumping ahead a little, I also value the clutch event differently for runs value since it is situational, but more on that later.

HR and Single park events- I convert these to singles, homeruns, and outs based on a 9/9 environment.

double plays- I convert ground ball a's assuming 20%, result in a double play

Stolen Bases- I am not going to fully describe this process, but it is based on OBP less HR's, less Xtra base hits. This result is then filtered through the lead probability's from the card and their safe chances to determine how often they will be sent, to arrive at how many SB chances a player will have. Depending on the steal chance, I multiply the SB chances times a success rate, 65% is the minimum I use, below that and I assume the runner will not attempt the steal. Note I currently do not penalize runners further for have caught stealing results while attempting the lead. There is obviously an impact for this but the effort verse the overall value wasn't worth the work to develop.

Average pitchers card

The key assumption here is that I assume 2 ballpark homeruns and 2 home run events on the pitchers card and a few OBP points and Total bases different depending on the hand of the pitcher

So lets stop their for a minute. All of this so far is just converting a players own card into Runs. Additionally I will ultimately add on:

Defense
Run rating (a function of run rating and OBP)
Team OBP (NERP concept)
Potential platoon buff if one side of a players card produces a run value below that of a $.50 cent player

I do not value hit and run or bunting into a players card

Mark

I originally put this in The Big Post for 2016, you can read more there if you like.

Even for a non-stat-nerd, that was fascinating. Thanks for (re)sharing.

I feel for Mark. It hasn't been that long to price the cards. But on the hand, this is a business. Not allocating sufficient resources to producing a product to the point that the product is delayed should get some push back from paying customers. Honestly, they could head most of that off by having better communication. Yes, there would still be complaints (it's the Internet after all), but if they said we're targeting 3/5/20, but ran into some issues that we want to make sure are taken care of prior to releasing the cards, I'd be 100% ok with it. Well, maybe 90% ok.

I normally only play a couple of seasonal teams to start with to get familiar with the cards, as I prefer the daily game. Play the seasonal in the off season. But it amazes me they can produce a game that updates on a daily basis yet somehow be delayed in one where the relevant info has been available for 6 months. Sounds like they really need someone who can write code

The problem isn’t developing the algorithms. That’s the easy part.

The problems arise from the very small imbalances peculiar to every card set. How do you price players who will only ever be used as a DH? How do you account for platoons? How do you account for the MLB players who accrue a lot of stats crushing lefties (the 9L guys)?

The necessary amount of tweaks are basically infinite, so I think a certain amount of understanding would go a long way here.

Mark is an extremely polite gentleman. He is also a manager just like all of us. If you have suggestions to make SOM Online pricing better, please contact him direct, I know he will be happy to hear from you. He wants the same great game we all want.

If you have suggestions for improving the SOM business model, send your input to John Garcia at the Strat help desk.

The problem isn’t developing the algorithms. That’s the easy part.

The problems arise from the very small imbalances peculiar to every card set. How do you price players who will only ever be used as a DH? How do you account for platoons? How do you account for the MLB players who accrue a lot of stats crushing lefties (the 9L guys)?

The necessary amount of tweaks are basically infinite, so I think a certain amount of understanding would go a long way here.

Mark is an extremely polite gentleman. He is also a manager just like all of us. If you have suggestions to make SOM Online pricing better, please contact him direct, I know he will be happy to hear from you. He wants the same great game we all want.

If you have suggestions for improving the SOM business model, send your input to John Garcia at the Strat help desk.

As is mentioned often, a little communication would be a huge improvement. I'm sure they are bombed with this as it's an issue year round. I also understand they most likely have limited staff and receive countless questions. But I also know businesses depend on their customers. I don't take threats of leaving seriously to be honest. But leaving customers in the dark so many times is not good business practice.

I for one don't expect perfection from them. If a card is a little off pricing wise it's not going to bother me. Everyone has the same opportunity to draft a bargain or ignore an over price. There are cards every year I target because I consider them a bargain, and ignore because I consider them not a good investment. Companies release things all the time then make adjustments. I have to think they make more money off the seasonal card sets than any other except the ATG series. Put it out and let folks play, with a disclaimer that in future leagues adjustments may be made. Make the money!

If you have suggestions for improving the SOM business model, send your input to John Garcia at the Strat help desk.

You mean the gentleman who hasn't posted on his own blog for almost TWO YEARS? Are we sure he still works at Strat-O-Matic?

Is Mark a paid employee of Strat-O-Matic, or does he voluntarily price the 365 card sets?

sociophil