Will agree with what others have said, especially about the midpoint perhaps being meaningless as it could be changed by a single bid or offer. I trade extensively in the SPX. Generally all the market makers will value certain spreads the same. These will include front week/month verticals, butterflies and the like. These are the easy ones for them to price. It gets more confusing as you move to the back months or look at spreads where they are simply buying or selling premium. Here their models will change quickly based on the order flow in the pit. If you are going to be entering front month verticals or similar hedged spreads, the slippage will be very small, generally about .05. So if you are selling a vertical that is "worth" .80, the market will usually be .75-.85. Depending on what's going on at the time or the price of the spread, the MM's may want .10 off of fair value, but I have found that .05 is usually enough to give up and get a fill. Unfortunately SPX spreads need to trade in at least .05 increments. The SPX is the toughest of the indexes for the retail investor to price as the markets are so wide. If each market is of similar width, generally the mid point is close fair value. I will generally look at the SPXPM to get more of an idea of where the SPX options should be as these markets are much tighter. The problem with the SPX is only one market maker sends quotes to the screen (unlike all other products). This is why they are so wide. Your best bet is once you determine the approximate fair value is send a 1 lot spread a little away from fair value (in your favor) and adjust the price until you are filled. If you are dealing with front month spreads that don't have too much premium risk (such as straddles), I would list the slippage to fair value to get filled as follows: SPX .05 VIX ..03 (these spreads can trade in pennies) SPY .02 (or less) OEX .10 XEO .10 Remember you can look at the COB on the CBOE web site and see what spreads are out there.
Again, you need to focus less on the midpoint. The spreads will trade off of fair value, which may or may not be the midpoint. I don't believe you can quantify which spreads will trade better or worse to fair value. This changes with order flow. If market makers have been buying premium they will be more likely to sell a spread the will hedge this position and require less value for it. Really no way to forecast this.
What type of model / software do market markers use to arrive at their own vol skew and fair values? For instance, I heard that the SABR volalility model is used to price options in the interest rate market. http://en.wikipedia.org/wiki/SABR_volatility_model How about in the index option market? What models are MM firms using to set their vol skews? How do MM arrive at their vol skews?