Can anyone point me to any research or information about the false negative rates of TMB? And has anyone studied what the probability is of a positive luminol hit and then a false negative for TMB?
I keep seeing the pro guilt folks arguing that blood comes up negative in a TMB test 50% of the time and can't imagine that is true?
That's a misinterpretation of what Gino said. She testified in her experience that due to the greater selectivity of TMB, that roughly half of all things picked up by luminol would test negative with a TMB test. In other words there are things other than blood that will cause a positive reaction to luminol and TMB, and in her experience there's about twice as many that do so with luminol than TMB as the latter is less likely to be 'fooled.'
As for false negatives, you won't find much literature about that because it's a natural chemical reaction. As long as there's something to oxidize the reagent TMB (Tetramethylbenzidine) then it will change colors, just like luminol, another reagent, will light up really cool, called 'chemiluminescence.' Think of it like water boiling, as long as there's enough heat, water will boil at some point, though it may be affected by air pressure or something in the water.
In the case of a TMB test the things that can interfere with it are basically dilution, age and the presence of a reduction agent. After the discovery of the negative TMB tests, Stefanoni hit the books and developed the 'hypothesis' that the material tested was between the detection threshold of luminol and TMB, meaning it was too diluted to be picked up with TMB but could be revealed with luminol. Unfortunately she, nor Massei, realized that doesn't work either, as if it was diluted to that extent it wouldn't have lit up so much for the pictures, and how she described in court that she 'knew' it was blood due to the intensity of the reaction.
As for age, that doesn't work either, they're talking on the order of
years not weeks. There's a number of papers that have been posted by both sides in this debate demonstrating positive blood tests after years of potential decay.
As for reduction agents, that ought to have interfered with the luminol test as well, and if for some esoteric reason they didn't, then it would have to have been extremely localized as it didn't interfere with the TMB tests performed on Rudy's bloody shoeprints right next to some of them, making it pretty much assured they would have had to have been actually on the feet (or whatever--some of these don't actually look much like footprints to me) of the person who made them, and a lot of the most common reduction agents are kinda nasty. One is aluminum-something or another which is rather acidic. At this point I opted out, any further questions can be directed to the
biochemist! 
Just so this might make more sense, here's how it's supposed to work. They spray luminol in suspected areas as it's easy to use on big areas and it will discover potential blood even when you can't see it. Then they look at what lights up and ones that look like they might be blood traces (a perfect square far more likely to be where a metallic object sat for years) and they test them with TMB to verify. Then they take the sample back to the lab for the
confirmation test which actually proves for
human blood, and a DNA test it to determine who it belonged to.
In this case Stefanoni applied the luminol, tested the items with TMB and got a negative, then took them back where (almost all) tested negative for DNA. Nothing was ever reported about a confirmatory test. So they went to court, lied about the negative TMB test, had it discovered later in the process and then came up with that 'hypothesis' that it could be between the detection thresholds and Comodi said 'It could be turnip juice or blood, you decide.' Then they pretended they hadn't lied about the TMB test because only a confirmatory test, that they never reported doing, actually proves for blood!
So in other words, they tried to claim because of the presumptive luminol test it was definitely blood, but they didn't lie about the TMB test because it's a presumptive test too and doesn't count as a blood test, so they never actually tested for blood because they didn't do (or report...) a confirmatory test!
'Oh what a tangled web we weave, when first we practice to deceive.'
. Actually, I've found your informed analysis and objectivity on this case hard to fault, and a real resource in understanding this case. If only certain others imagining themselves to be authorities on the details of the case could be as impartial.
