Occam’s razor is a principle some people (such as biographer W.T. Schultz) have cited to explain Elliott Smith’s death: you know, the simplest explanation is usually the right one…. Depression, self-hatred, and addiction lead to suicide and there’s no reason to be looking for a more complex explanation. Really?
The concept, named after 14th-century philosopher William of Occam, is highly disputable in cases like this one because they are not simple equations, and investigations have little to do with elegant mathematical demonstrations. The principle has not always served science for the best (just read this article in the Atlantic), and there’s no way I can see good use of Occam’s razor in any science dealing with life and psychology. Why? Because life, in general, is messy and chaotic, because nothing is simple and nature isn’t usually parsimonious. Study any life system and you will always encounter complexity and counter-intuitive explanation. Elliott’s case involves forensic science and psychology, and none of these disciplines is simple, because they must take into consideration many parameters. Evoking Occam’s razor, in this case, seems to be like glossing many facts under an academic expression. Occam’s razor would only apply if the second theory (the non-suicide theory) would imply making a number of assumptions… However, there is no need to make assumptions, only a series of facts which challenge the simplest and most superficial hypothesis.
I have already discussed the forensic parameters of the case, but forensic literature actually offers a large amount of data related to many of these details, and it’s tempting to use them to estimate the probability that Elliott actually committed suicide. Let’s review the details of the case: The lack of hesitation marks surrounding the stab wounds, the stabbing through the clothes and the bones injuries: Elliott’s sternum was injured as well as his rib cartilage…. I have not considered a few other facts because, either they were not strong indicators of the method of death, or they were not clear enough. For example, Elliott had small fresh wounds on his upper right arm and palm, which could have been defense injuries. However, it was not proven, as the coroner could not determine for sure if they were real defense wounds. Elliott also had no illegal drug or alcohol in his system, but once again, this doesn’t seem to be a strong indicator to decide between suicide and murder… There is the last item I didn’t consider, the direction of the stab wounds, mostly because the autopsy report says ‘slightly downward’ and not vertical. Many studies indicate that this also could be an important parameter to determine the difference between a suicide and a homicide: ‘the horizontal orientation is assumed to be more common in suicides than in homicides because of the manner of holding the knife, which results in the victim’s self-inflicted injuries being more often horizontal’ . However more studies are needed and without more details (X rays?) about Elliott’s case, it is difficult to decide.
But let’s concentrate on what we know for sure:
* Suicide by stabbing only represents 1.6-3% of all suicides. You can find numbers in many articles related to forensic literature: ‘Actually sharp force is a rare method of suicide constituting only 2–3% of all suicides in various countries’ , ‘Suicide by stabbing is uncommon, constituting only 1.6-3.0% of suicide attempts’ , Suicide by sharp force injuries accounted for 2.5% of all suicides in our prefecture during 1995–2005’ , ‘Fatalities from sharp force are mainly homicidal as suicides by sharp force represent only 1.6-3% of all suicides’ .
Although the estimation varies a little bit, we can consider the average of this range, that is 2.3%.
* Stabbing in the chest represents only a certain percentage of this type of suicides, as many people stab themselves in the neck or the abdomen, or even injure other parts of the body: ‘The most common injury sites were the abdomen (46%), neck (33%), and chest (20%)’ . And if you consider suicidal injuries in general, the numbers can be a bit higher, like in this extensive study which took in consideration 595 potentially relevant articles: ‘The thorax is identified as the most common site in homicide, particularly when the injury is single’…’The present review of the literature confirms that thorax is the most frequent region involved in both manners of death, albeit with some differences. In fact, the thorax is targeted in more than half (54%) of homicidal wounds but only in about a third (36.6%) of suicidal ones’ .
Again let’s take the average of these numbers (20 % and 36.6%), that is 28.3%.
* Hesitation or tentative marks are observed in most cases of suicide by stabbing. Obviously statistics vary in the articles I have consulted, but data are certainly not in contradiction with each other: ‘Tentative marks were present in 18 cases (64%) and these injuries were usually in close proximity to the fatal wound’ , ‘In 27 of 37 cases [73%], hesitation marks were observed in close proximity’ , ‘Classical indicators of suicidal intent, for example, suicide notes and the presence of hesitation injuries, were found in 28% and 80%, respectively’ , ‘Defense wounds were reported in 49% of the homicides and “hesitation” wounds were found in 65% of the suicides’, ‘Tentative marks were present in 50 (77%) fatalities and the number varied from 1 to 60 per case’ , ‘Tentative wounds were found in eight cases altogether (50%) and hesitation marks in 6 cases (37.5%). Nevertheless, both or at least one of these lesions were present in 14 cases (87.5%). Such hesitation marks are shallow lesions, parallel to and in close proximity of fatal wounds, and are considered to be, along with tentative (non-fatal) wounds, the most useful indicator to distinguish suicide from homicide – over 70%, in most studies’ , ‘A total of 58 suicides (7 women; 51 men) and 149 homicides (59 women; 60 men) were found, of which 74% (n = 43) were positive for hesitation marks and 61% (n = 91) for defense wounds’ ‘Hesitation marks were present in 322/672 cases (47.9%) and absent in 194 cases (28.9%) and information was not available in 156 cases (23.2%). Excluding these 156 cases, hesitation injuries were present in 62% of the cases, agreeing with the percentage already reported in literature’ 
If we average all these numbers (64%, 73%, 80%, 65%, 77%, 87.5%, 74%, and 62%) we can safely say that hesitation/tentative marks are present in the large majority of suicide (average of 72.8%), so Elliott’s case belongs to the 27.1% who did not present any hesitation marks.
* Studies about suicide by stabbing point out that clothes are generally not damaged because they are removed before the act. Although numbers vary greatly, a few studies bring more indications: ‘A preference for certain anatomical locations (throat, precordium, epigastrium, wrists) was confirmed as was the tendency to expose the skin before inflicting suicidal wounds’ , ‘In 16 (57%) cases stab wounds occurred after clothing had been purposely moved aside but wounds occurred through clothing in 8 cases (28%)’ , ‘Of 28 cases with fatal cut or stab wounds localized in the trunk, 11 cases (39%) had clothing damage’ , ‘Perforation of clothing was present in 16 (52%) out of 31 stab injuries to the trunk’ , ‘In suicides, sharp wounds are usually sustained at sites not covered by clothing or at sites exposed after the clothing is pulled up. This was the case in 8 of the victims in our study, being that in 6 cases the wounded areas were already exposed. The clothing was damaged by the implement in only 2 cases (12.5%), both with stab injuries to the chest (the victims suffered from schizophrenia)’ , ‘In the case of suicide, clothing injuries were visible in 72/672 cases (10.7%) and absent in 235 cases (35%) and information was not available in 365 cases (54.3%). Excluding these 365 cases, clothing was damaged in 23% of cases and not damaged in 77% of cases’, .
Once again we can average these numbers, and thus suicides by stabbing presenting damage in clothes (28%, 39%, 52%, 12.5%, 23%) only represent an average of 30.9% of all cases.
* In an extensive study about stabbing (murder versus suicide), bones injuries were found to be a strong indication of homicide: ‘The presence of bone or cartilage wounds was predictive of homicide and their absence was predictive of a suicide.’… ‘In homicides, bone or cartilage wounds were present in 52 cases (74.3%), absent in 17 cases (24.3%) and not reported in 1 case (1.4%). In suicides, bone or cartilage wounds were present in 7 cases (14.6%) and absent in 41 cases (85.4%). Bone or cartilage wounds were thus found to be a significant predictive factor relative to the manner of death’ 
As Elliott had sternum and cartilage injuries, his case, if considered a suicide, would fall in this 14.6%.
Thus, it is simple to compile all these data: 2.3% (percentage of suicide by stabbing) x 28.3% (percentage of stabbing in the chest) x 27.1% (percentage of suicide with no hesitation marks) x 30.9% (percentage of suicide by stabbing presenting damage in clothes) x 14.6% (percentage of suicide by stabbing presenting bone injuries)… And the result is astronomically low, approximately 0,008%.
Based on this research, there is a 0.008% probability that Elliott committed suicide, which makes his case an extremely rare case in forensic literature. Unheard of.
I was told that ‘the devil is in the details’, and that’s a good thing I don’t believe in the devil… actually, the truth is in the details.
 Non-fatal suicide attempt by intentional stab wound: Clinical management, psychiatric assessment, and multidisciplinary considerations: James M Badger, Shea C Gregg, and Charles A Adams, Jr, J Emerg Trauma Shock, v.5(3); Jul-Sep 2012
 Retrospective study on suicidal cases by sharp force injuries: Setsuko Fukube, Takahito Hayashi, MD, Yuko Ishida, PhD, Hitoshi Kamon, Mariko Kawaguchi, Akihiko Kimura, PhD, Toshikazu Kondo, PhD, MD, An International Journal of Forensic and Legal Medicine, April 2008 Volume 15, Issue 3, Pages 163–167
 Patterns in sharp force fatalities–a comprehensive forensic medical study: Part 2. Suicidal sharp force injury in the Stockholm area 1972-1984: Karlsson T, Ormstad K, Rajs J. J Forensic Sci.1988 Mar; 33(2):448-61.
 Suicidal and homicidal sharp force injury: a 5-year retrospective comparative study of hesitation marks and defense wounds. Stephanie R Racette, Célia Kremer, Anne Desjarlais, Anny Sauvageau 2008in Forensic science, medicine, and pathology
 Homicidal and suicidal sharp force fatalities: Autopsy parameters in relation to the manner of death: Christophe Brunel, Christophe Fermanian, Michel Durigon, Geoffroy Lorin de la Grandmaison, Forensic Science International, Volume 198, Issues 1–3, 20 May 2010, Pages 150-154
 Suicidal or Homicidal Sharp Force Injuries? A Review and Critical Analysis of the Heterogeneity in the Forensic Literature: Fabio De‐Giorgio M.D., Ph.D., Maria Lodise M.D., Gianluigi Quaranta M.D., Ph.D., Antonio G. Spagnolo M.D., Ernesto d’Aloja M.D., Ph.D., Vincenzo L. Pascali M.D., Ph.D., Vincenzo M. Grassi M.D. J Forensic Sci.2015 Jan;60 Suppl 1:S97-107. doi: 10.1111/1556-4029.12673. Epub 2014 Nov 27.